Leaderboard

Best benchmark result for each model + quantization + GPU combination, ranked by throughput. Data refreshes automatically.

Connecting to data source…

Model	Quant	GPU	Context	Throughput (tok/s)↓	Avg TTFT (ms)	P99 TTFT (ms)	Avg ITL (ms)	P99 ITL (ms)	Avg W	tok/W
gpt-oss-20b-Q4_1	Q4_1	NVIDIA GeForce RTX 5090	32,768	1491.1	199.8	349.3	20.74	22.13	192	7.75
gpt-oss-20b-Q5_K_M	Q5_K_M	NVIDIA GeForce RTX 5090	130,064	1421.2	199.5	349.6	21.81	23.04	213	6.67
gpt-oss-20b-Q4_K_S	Q4_K_S	NVIDIA GeForce RTX 5090	32,768	1374.2	214.4	369.6	22.44	35.08	217	6.32
gpt-oss-20b-Q3_K_M	Q3_K_M	NVIDIA GeForce RTX 5090	130,064	1370.1	338.1	1079.6	22.02	37.27	202	6.77
gpt-oss-20b-UD-Q6_K_XL	Q6_K_XL	NVIDIA GeForce RTX 5090	16,384	1369.4	208.4	357.7	22.51	35.92	215	6.37
gpt-oss-20b-Q5_K_S	Q5_K_S	NVIDIA GeForce RTX 5090	16,384	1367.4	325.9	1200.9	22.16	29.70	164	8.34
gpt-oss-20b-Q2_K	Q2_K	NVIDIA GeForce RTX 5090	32,768	1367.1	207.3	347.1	22.56	28.54	212	6.45
gpt-oss-20b-Q3_K_S	Q3_K_S	NVIDIA GeForce RTX 5090	16,384	1361.4	327.1	1081.5	22.10	36.33	203	6.69
gpt-oss-20b-UD-Q8_K_XL	Q8_K_XL	NVIDIA GeForce RTX 5090	130,064	1358.9	223.0	387.9	22.70	35.41	215	6.31
gpt-oss-20b-Q4_K_M	Q4_K_M	NVIDIA GeForce RTX 5090	32,768	1357.8	207.6	340.6	22.70	34.92	217	6.27
gpt-oss-20b-F16	F16	NVIDIA GeForce RTX 5090	16,384	1355.0	212.1	348.4	22.73	34.78	223	6.08
gpt-oss-20b-UD-Q4_K_XL	Q4_K_XL	NVIDIA GeForce RTX 5090	16,384	1344.6	220.2	352.9	22.90	35.20	214	6.27
gpt-oss-20b-Q6_K	Q6_K	NVIDIA GeForce RTX 5090	32,768	1338.6	220.9	386.9	22.91	44.13	217	6.17
gemma-4-E2B-it-Q8_0	Q8_0	NVIDIA GeForce RTX 5090	32,768	1334.7	101.3	184.4	20.96	56.79	189	7.07
gpt-oss-20b-Q4_0	Q4_0	NVIDIA GeForce RTX 5090	130,064	1334.3	222.4	385.5	22.98	44.37	214	6.24
gemma-4-E2B-it-Q4_1	Q4_1	NVIDIA GeForce RTX 5090	32,768	1326.1	101.1	186.6	20.99	50.08	175	7.56
gpt-oss-20b-Q8_0	Q8_0	NVIDIA GeForce RTX 5090	32,768	1322.6	249.4	480.0	23.16	45.37	219	6.04
gpt-oss-20b-Q2_K_L	Q2_K_L	NVIDIA GeForce RTX 5090	16,384	1316.7	247.4	423.4	23.09	41.03	219	6.02
gemma-4-E2B-it-Q5_K_S	Q5_K_S	NVIDIA GeForce RTX 5090	32,768	1311.5	97.7	187.9	21.29	52.30	195	6.74
gemma-4-E2B-it-UD-Q2_K_XL	Q2_K_XL	NVIDIA GeForce RTX 5090	130,064	1308.0	129.9	355.2	20.91	55.61	147	8.88
gemma-4-E2B-it-Q5_K_M	Q5_K_M	NVIDIA GeForce RTX 5090	130,064	1307.8	100.8	185.8	21.23	54.57	190	6.88
gemma-4-E2B-it-IQ4_XS	IQ4_XS	NVIDIA GeForce RTX 5090	16,384	1302.3	102.8	194.7	20.84	50.44	175	7.45
gemma-4-E2B-it-Q4_K_S	Q4_K_S	NVIDIA GeForce RTX 5090	32,768	1290.1	122.3	299.6	21.00	65.33	148	8.73
gemma-4-E2B-it-UD-Q5_K_XL	Q5_K_XL	NVIDIA GeForce RTX 5090	130,064	1283.1	99.3	189.0	21.45	50.89	197	6.53
gemma-4-E2B-it-Q6_K	Q6_K	NVIDIA GeForce RTX 5090	16,384	1280.2	107.1	191.1	21.33	55.41	201	6.38
gemma-4-E2B-it-Q3_K_S	Q3_K_S	NVIDIA GeForce RTX 5090	130,064	1270.4	94.5	206.9	21.26	48.25	189	6.74
gemma-4-E2B-it-UD-Q8_K_XL	Q8_K_XL	NVIDIA GeForce RTX 5090	16,384	1265.5	114.0	339.3	21.65	54.16	156	8.11
gemma-4-E2B-it-UD-Q4_K_XL	Q4_K_XL	NVIDIA GeForce RTX 5090	130,064	1262.6	102.8	200.0	21.83	53.13	191	6.60
gemma-4-E2B-it-Q3_K_M	Q3_K_M	NVIDIA GeForce RTX 5090	130,064	1246.8	115.8	312.8	22.08	59.39	154	8.10
gemma-4-E2B-it-UD-Q6_K_XL	Q6_K_XL	NVIDIA GeForce RTX 5090	130,064	1241.8	101.5	188.0	21.48	51.09	200	6.22
gemma-4-E2B-it-Q4_0	Q4_0	NVIDIA GeForce RTX 5090	16,384	1241.7	110.8	306.5	21.84	52.55	142	8.76
gemma-4-E2B-it-Q4_K_M	Q4_K_M	NVIDIA GeForce RTX 5090	32,768	1233.9	105.5	191.3	20.98	57.08	196	6.30
gemma-4-E4B-it-UD-IQ2_M	IQ2_M	NVIDIA GeForce RTX 5090	32,768	1227.6	230.7	387.6	23.34	49.42	170	7.24
gemma-4-E2B-it-IQ4_NL	IQ4_NL	NVIDIA GeForce RTX 5090	130,064	1192.7	100.8	186.0	21.48	51.99	180	6.62
gemma-4-E2B-it-BF16	BF16	NVIDIA GeForce RTX 5090	16,384	1188.0	114.7	239.0	22.44	59.66	176	6.73
gemma-4-E2B-it-UD-IQ2_M	IQ2_M	NVIDIA GeForce RTX 5090	130,064	1169.9	97.6	204.4	23.35	47.79	183	6.39
gemma-4-E2B-it-UD-IQ3_XXS	IQ3_XXS	NVIDIA GeForce RTX 5090	32,768	1161.6	93.6	178.9	22.30	48.50	186	6.24
gemma-4-E2B-it-UD-Q3_K_XL	Q3_K_XL	NVIDIA GeForce RTX 5090	32,768	1147.2	99.5	187.7	22.48	53.00	195	5.89
gemma-4-E4B-it-Q4_1	Q4_1	NVIDIA GeForce RTX 5090	130,064	1135.2	154.3	401.7	25.13	61.84	167	6.81
gemma-4-E4B-it-UD-IQ3_XXS	IQ3_XXS	NVIDIA GeForce RTX 5090	32,768	1131.2	149.9	386.8	24.85	57.60	171	6.61
gemma-4-E4B-it-UD-Q2_K_XL	Q2_K_XL	NVIDIA GeForce RTX 5090	16,384	1118.0	283.1	848.7	24.69	63.96	149	7.52
gemma-4-E4B-it-Q4_0	Q4_0	NVIDIA GeForce RTX 5090	130,064	1106.7	173.0	373.7	25.51	69.75	144	7.69
gemma-4-E4B-it-IQ4_XS	IQ4_XS	NVIDIA GeForce RTX 5090	16,384	1103.5	137.8	376.3	25.13	55.64	222	4.96
gemma-4-E4B-it-Q5_K_S	Q5_K_S	NVIDIA GeForce RTX 5090	130,064	1091.2	155.9	327.7	26.46	81.92	156	7.01
gemma-4-E4B-it-UD-Q4_K_XL	Q4_K_XL	NVIDIA GeForce RTX 5090	16,384	1082.1	150.4	398.4	25.94	63.99	185	5.87
gemma-4-E4B-it-UD-Q5_K_XL	Q5_K_XL	NVIDIA GeForce RTX 5090	16,384	1050.8	159.1	394.8	26.97	69.29	194	5.43
gemma-4-E4B-it-IQ4_NL	IQ4_NL	NVIDIA GeForce RTX 5090	16,384	1047.7	147.6	395.5	26.28	62.79	222	4.73
gemma-4-E4B-it-Q8_0	Q8_0	NVIDIA GeForce RTX 5090	32,768	1046.6	154.9	383.2	27.15	67.36	188	5.58
gemma-4-E4B-it-Q5_K_M	Q5_K_M	NVIDIA GeForce RTX 5090	130,064	1044.9	162.3	403.8	26.89	71.90	194	5.39
gemma-4-E4B-it-UD-Q6_K_XL	Q6_K_XL	NVIDIA GeForce RTX 5090	32,768	1036.2	161.3	417.4	27.43	68.97	195	5.32
gemma-4-E4B-it-Q4_K_S	Q4_K_S	NVIDIA GeForce RTX 5090	130,064	1020.0	159.6	402.1	26.73	72.77	189	5.40
gemma-4-E4B-it-Q6_K	Q6_K	NVIDIA GeForce RTX 5090	32,768	998.8	157.6	410.0	27.70	68.97	202	4.95
gemma-4-E4B-it-UD-Q3_K_XL	Q3_K_XL	NVIDIA GeForce RTX 5090	130,064	997.2	162.5	381.8	27.61	78.41	159	6.27
gemma-4-E4B-it-Q4_K_M	Q4_K_M	NVIDIA GeForce RTX 5090	32,768	996.7	157.4	434.5	28.28	71.55	159	6.28
gemma-4-E4B-it-UD-Q8_K_XL	Q8_K_XL	NVIDIA GeForce RTX 5090	32,768	990.1	159.0	387.1	28.07	67.52	197	5.04
gemma-4-E4B-it-Q3_K_S	Q3_K_S	NVIDIA GeForce RTX 5090	32,768	986.2	191.9	469.8	27.52	66.89	191	5.17
gemma-4-E4B-it-Q3_K_M	Q3_K_M	NVIDIA GeForce RTX 5090	32,768	980.9	163.3	397.1	27.22	65.41	253	3.88
gemma-4-26B-A4B-it-UD-IQ2_XXS	IQ2_XXS	NVIDIA GeForce RTX 5090	32,768	928.9	804.8	2379.4	30.94	88.81	225	4.14
gemma-4-E4B-it-BF16	BF16	NVIDIA GeForce RTX 5090	130,064	898.3	170.7	410.9	31.83	91.81	231	3.88
gemma-4-26B-A4B-it-UD-IQ2_M	IQ2_M	NVIDIA GeForce RTX 5090	32,768	897.1	992.1	2482.4	31.25	80.27	264	3.40
gemma-4-26B-A4B-it-MXFP4_MOE		NVIDIA GeForce RTX 5090	130,064	844.9	520.4	1694.2	34.37	134.19	231	3.66
gemma-4-26B-A4B-it-UD-Q3_K_M	Q3_K_M	NVIDIA GeForce RTX 5090	65,536	816.6	668.7	2136.1	35.11	125.15	236	3.46
gemma-4-26B-A4B-it-UD-Q4_K_M	Q4_K_M	NVIDIA GeForce RTX 5090	32,768	815.9	431.5	1280.4	35.92	140.31	238	3.43
gemma-4-26B-A4B-it-UD-Q4_K_XL	Q4_K_XL	NVIDIA GeForce RTX 5090	65,536	811.5	536.6	1284.4	36.01	138.95	275	2.95
gemma-4-26B-A4B-it-UD-Q3_K_XL	Q3_K_XL	NVIDIA GeForce RTX 5090	32,768	810.6	726.6	1767.9	35.18	102.75	269	3.02
gemma-4-26B-A4B-it-UD-Q4_K_S	Q4_K_S	NVIDIA GeForce RTX 5090	65,536	810.5	486.3	1314.8	36.24	143.94	275	2.94
Qwen3.5-0.8B-Q4_1	Q4_1	NVIDIA GeForce RTX 5090	32,768	808.5	117.2	158.0	4.50	6.40	—	—
gemma-4-26B-A4B-it-UD-Q2_K_XL	Q2_K_XL	NVIDIA GeForce RTX 5090	65,536	807.4	638.1	2023.6	35.31	135.42	237	3.41
gemma-4-26B-A4B-it-UD-IQ3_S	IQ3_S	NVIDIA GeForce RTX 5090	65,536	795.9	580.9	1900.0	36.27	130.75	241	3.31
Qwen3.5-0.8B-Q4_0	Q4_0	NVIDIA GeForce RTX 5090	130,064	788.6	84.9	122.3	4.65	6.50	158	4.99
gemma-4-26B-A4B-it-UD-IQ4_NL	IQ4_NL	NVIDIA GeForce RTX 5090	32,768	786.9	306.2	1004.7	37.95	157.86	239	3.30
Qwen3.5-0.8B-IQ4_NL	IQ4_NL	NVIDIA GeForce RTX 5090	8,192	786.2	1205.5	1398.0	4.82	6.59	169	4.65
Qwen3.5-0.8B-Q8_0	Q8_0	NVIDIA GeForce RTX 5090	16,384	777.3	1285.5	1435.0	4.75	6.50	174	4.48
gemma-4-26B-A4B-it-UD-IQ4_XS	IQ4_XS	NVIDIA GeForce RTX 5090	32,768	776.6	310.7	1007.3	38.51	144.25	241	3.23
Qwen3.5-0.8B-Q6_K	Q6_K	NVIDIA GeForce RTX 5090	130,064	775.4	113.9	159.9	4.63	6.51	—	—
Qwen3.5-0.8B-UD-Q2_K_XL	Q2_K_XL	NVIDIA GeForce RTX 5090	130,064	775.1	1367.3	1470.5	4.73	6.62	172	4.51
Qwen3.5-0.8B-IQ4_XS	IQ4_XS	NVIDIA GeForce RTX 5090	16,384	774.5	1299.6	1433.0	4.81	6.57	—	—
Qwen3.5-0.8B-UD-IQ2_XXS	IQ2_XXS	NVIDIA GeForce RTX 5090	16,384	773.7	1369.5	1473.9	4.75	6.55	—	—
Qwen3.5-0.8B-IQ4_NL	IQ4_NL		32,768	768.8	200.1	547.1	39.86	59.03	—	—
Qwen3.5-0.8B-Q3_K_S	Q3_K_S	NVIDIA GeForce RTX 5090	8,192	766.9	115.0	158.1	4.78	6.68	170	4.51
gemma-4-26B-A4B-it-UD-IQ3_XXS	IQ3_XXS	NVIDIA GeForce RTX 5090	65,536	766.2	533.0	1636.3	37.64	141.80	241	3.18
gemma-4-26B-A4B-it-UD-Q5_K_S	Q5_K_S	NVIDIA GeForce RTX 5090	32,768	763.6	395.5	1364.9	37.71	141.74	243	3.14
Qwen3.5-2B-IQ4_NL	IQ4_NL	NVIDIA GeForce RTX 5090	65,551	759.5	118.3	163.5	4.82	6.70	246	3.09
Qwen3.5-0.8B-Q3_K_M	Q3_K_M	NVIDIA GeForce RTX 5090	8,192	752.6	115.7	156.9	4.88	6.78	171	4.41
Qwen3.5-0.8B-UD-IQ2_M	IQ2_M	NVIDIA GeForce RTX 5090	130,064	746.2	1323.8	1473.9	5.01	6.72	—	—
Qwen3.5-2B-Q4_0	Q4_0	NVIDIA GeForce RTX 5090	32,768	744.2	1369.1	1499.4	5.03	6.73	205	3.63
Qwen3.5-0.8B-UD-Q8_K_XL	Q8_K_XL	NVIDIA GeForce RTX 5090	8,192	741.1	115.3	159.0	4.96	6.78	168	4.41
gemma-4-26B-A4B-it-UD-Q5_K_XL	Q5_K_XL	NVIDIA GeForce RTX 5090	32,768	739.9	419.9	1133.4	38.26	150.63	287	2.58
Qwen3.5-0.8B-UD-Q6_K_XL	Q6_K_XL	NVIDIA GeForce RTX 5090	130,064	739.8	1306.4	1486.4	5.14	7.50	171	4.34
Qwen3.5-0.8B-UD-Q5_K_XL	Q5_K_XL	NVIDIA GeForce RTX 5090	130,064	737.3	116.7	155.5	4.99	6.85	166	4.44
gemma-4-26B-A4B-it-UD-Q5_K_M	Q5_K_M	NVIDIA GeForce RTX 5090	32,768	732.3	479.6	1108.4	37.90	148.31	284	2.58
Qwen3.5-0.8B-Q5_K_M	Q5_K_M	NVIDIA GeForce RTX 5090	16,384	732.2	85.1	123.9	5.08	6.87	174	4.21
Qwen3.5-0.8B-BF16	BF16	NVIDIA GeForce RTX 5090	130,064	730.0	116.4	155.8	5.04	6.82	—	—
Qwen3.5-2B-Q4_1	Q4_1	NVIDIA GeForce RTX 5090	130,064	724.2	3886.3	4373.9	5.31	29.02	234	3.09
Qwen3.5-0.8B-Q4_K_S	Q4_K_S	NVIDIA GeForce RTX 5090	8,192	723.5	113.5	158.8	5.03	6.95	—	—
Qwen3.5-2B-UD-Q2_K_XL	Q2_K_XL	NVIDIA GeForce RTX 5090	65,551	714.0	121.8	171.9	5.13	7.06	250	2.86
Qwen3.5-0.8B-UD-IQ3_XXS	IQ3_XXS	NVIDIA GeForce RTX 5090	8,192	713.5	116.3	153.8	5.17	7.06	166	4.30
Qwen3.5-0.8B-UD-Q3_K_XL	Q3_K_XL	NVIDIA GeForce RTX 5090	16,384	713.4	1402.7	1544.8	5.23	6.97	175	4.07
Qwen3.5-0.8B-Q4_K_M	Q4_K_M	NVIDIA GeForce RTX 5090	8,192	712.4	1345.1	1517.7	5.40	23.21	—	—
Qwen3.5-0.8B-Q5_K_S	Q5_K_S	NVIDIA GeForce RTX 5090	16,384	708.9	1336.7	1559.6	5.36	7.13	183	3.88
Qwen3.5-2B-Q4_0	Q4_0	NVIDIA GeForce RTX 5060 Ti	32,768	705.7	253.0	683.7	43.27	65.93	83	8.49
Qwen3.5-2B-Q3_K_M	Q3_K_M	NVIDIA GeForce RTX 5060 Ti	130,064	699.5	296.5	775.1	43.88	65.03	84	8.36
Qwen3.5-2B-Q3_K_S	Q3_K_S	NVIDIA GeForce RTX 5060 Ti	32,768	699.4	295.9	772.3	43.79	74.26	84	8.34
Qwen3.5-2B-Q4_1	Q4_1	NVIDIA GeForce RTX 5060 Ti	32,768	694.7	264.5	755.4	43.74	77.70	84	8.30
Qwen3.5-0.8B-UD-Q4_K_XL	Q4_K_XL	NVIDIA GeForce RTX 5090	130,064	694.2	1338.0	1576.5	5.42	8.81	—	—
Qwen3.5-2B-Q6_K	Q6_K	NVIDIA GeForce RTX 5090	16,384	690.9	117.0	169.9	5.28	9.64	—	—
Qwen3.5-2B-IQ4_XS	IQ4_XS	NVIDIA GeForce RTX 5090	65,551	690.1	1424.7	1599.3	5.58	25.90	162	4.26
Qwen3.5-2B-Q4_K_S	Q4_K_S	NVIDIA GeForce RTX 5060 Ti	32,768	690.0	229.3	737.7	44.11	70.40	85	8.12
Qwen3.5-0.8B-BF16	BF16		130,064	688.2	212.9	539.7	43.25	71.46	—	—
Qwen3.5-2B-IQ4_NL	IQ4_NL	NVIDIA GeForce RTX 5060 Ti	16,384	688.0	236.4	698.0	43.88	79.79	86	8.02
Qwen3.5-2B-IQ4_XS	IQ4_XS	NVIDIA GeForce RTX 5060 Ti	32,768	686.8	254.7	764.5	43.89	73.86	85	8.11
Qwen3.5-2B-Q3_K_S	Q3_K_S	NVIDIA GeForce RTX 5090	65,551	686.7	120.8	170.4	5.37	7.21	277	2.48
Qwen3.5-2B-IQ4_NL	IQ4_NL		16,384	682.6	259.9	822.2	44.33	88.12	—	—
Qwen3.5-2B-UD-IQ2_XXS	IQ2_XXS	NVIDIA GeForce RTX 5090	16,384	679.8	1547.1	1663.8	5.45	7.35	227	3.00
Qwen3.5-2B-Q8_0	Q8_0	NVIDIA GeForce RTX 5090	130,064	678.0	1451.4	1624.2	5.67	15.28	187	3.63
Qwen3.5-2B-UD-Q5_K_XL	Q5_K_XL	NVIDIA GeForce RTX 5060 Ti	16,384	677.7	264.6	758.6	44.74	79.67	87	7.79
Qwen3.5-2B-UD-Q4_K_XL	Q4_K_XL	NVIDIA GeForce RTX 5060 Ti	130,064	673.7	223.2	747.0	44.83	74.28	87	7.78
Qwen3.5-2B-Q4_K_M	Q4_K_M	NVIDIA GeForce RTX 5060 Ti	130,064	672.4	225.5	737.7	45.38	84.46	76	8.82
Qwen3.5-2B-Q5_K_S	Q5_K_S	NVIDIA GeForce RTX 5060 Ti	32,768	671.6	243.7	740.4	45.19	78.06	88	7.65
Qwen3.5-2B-Q6_K	Q6_K	NVIDIA GeForce RTX 5060 Ti	32,768	667.9	257.5	785.8	45.53	81.73	88	7.63
Qwen3.5-2B-UD-IQ2_M	IQ2_M	NVIDIA GeForce RTX 5090	65,551	664.4	1581.3	1700.3	5.58	7.51	233	2.86
Qwen3.5-2B-Q5_K_M	Q5_K_M	NVIDIA GeForce RTX 5060 Ti	16,384	663.6	232.4	792.2	45.53	82.61	90	7.41
Qwen3.5-2B-Q3_K_M	Q3_K_M	NVIDIA GeForce RTX 5090	32,768	657.6	116.5	167.1	5.57	7.33	279	2.35
Qwen3.5-2B-Q8_0	Q8_0	NVIDIA GeForce RTX 5060 Ti	16,384	649.8	229.8	756.5	46.88	96.73	74	8.77
Qwen3.5-2B-UD-Q6_K_XL	Q6_K_XL	NVIDIA GeForce RTX 5090	8,192	643.3	10282.8	11378.5	6.04	30.12	270	2.38
gemma-4-26B-A4B-it-UD-Q6_K_XL	Q6_K_XL	NVIDIA GeForce RTX 5090	8,192	634.9	211.8	517.9	23.46	92.57	237	2.67
Qwen3.5-2B-UD-IQ3_XXS	IQ3_XXS	NVIDIA GeForce RTX 5090	16,384	631.9	120.6	162.6	5.88	7.79	255	2.48
Qwen3.5-2B-UD-Q3_K_XL	Q3_K_XL	NVIDIA GeForce RTX 5090	65,551	629.7	1581.5	1732.8	6.08	7.78	222	2.83
DeepSeek-R1-Distill-Qwen-32B-Q2_K_L	Q2_K_L	NVIDIA GeForce RTX 5090	16,384	628.0	1036.0	3426.8	42.56	142.11	360	1.75
DeepSeek-R1-Distill-Qwen-32B-Q2_K	Q2_K	NVIDIA GeForce RTX 5090	16,384	627.4	1182.0	3372.2	42.56	146.71	408	1.54
Qwen3.5-2B-UD-Q6_K_XL	Q6_K_XL	NVIDIA GeForce RTX 5060 Ti	16,384	625.9	278.3	845.6	47.90	88.11	88	7.10
gemma-4-26B-A4B-it-UD-Q6_K	Q6_K	NVIDIA GeForce RTX 5090	8,192	624.9	205.0	482.9	23.94	97.48	237	2.64
Qwen3.5-2B-Q5_K_M	Q5_K_M	NVIDIA GeForce RTX 5090	16,384	622.1	1530.0	1756.1	6.16	27.22	227	2.74
gemma-4-26B-A4B-it-UD-Q8_K_XL	Q8_K_XL	NVIDIA GeForce RTX 5090	8,192	616.4	263.3	661.0	23.85	86.61	302	2.04
gemma-4-26B-A4B-it-Q8_0	Q8_0	NVIDIA GeForce RTX 5090	8,192	615.6	270.7	659.6	24.04	101.78	297	2.07
Qwen3.5-2B-UD-Q5_K_XL	Q5_K_XL	NVIDIA GeForce RTX 5090	8,192	610.9	91.2	132.7	6.12	7.97	260	2.35
Qwen3.5-2B-Q4_K_S	Q4_K_S	NVIDIA GeForce RTX 5090	130,064	608.8	4637.7	5177.3	6.37	33.27	265	2.30
Qwen3.5-2B-Q4_K_M	Q4_K_M	NVIDIA GeForce RTX 5090	8,192	608.3	1579.2	1791.6	6.37	29.63	231	2.64
gemma-4-E2B-it-BF16	BF16	NVIDIA GeForce RTX 5060 Ti	130,064	607.4	174.1	329.3	43.50	114.19	72	8.42
Qwen3.5-2B-Q5_K_S	Q5_K_S	NVIDIA GeForce RTX 5090	32,768	606.8	4725.5	5188.9	6.34	8.04	—	—
Qwen3.5-2B-UD-Q8_K_XL	Q8_K_XL	NVIDIA GeForce RTX 5090	32,768	597.3	117.8	167.2	6.26	8.07	—	—
Qwen3.5-2B-UD-Q4_K_XL	Q4_K_XL	NVIDIA GeForce RTX 5090	16,384	596.2	113.9	164.3	6.19	8.05	264	2.26
Qwen3.5-2B-UD-Q8_K_XL	Q8_K_XL	NVIDIA GeForce RTX 5060 Ti	130,064	590.0	283.7	860.2	50.70	90.05	87	6.78
Qwen3.5-2B-UD-IQ2_XXS	IQ2_XXS	NVIDIA GeForce RTX 5060 Ti	130,064	569.7	181.0	439.0	47.31	84.08	80	7.17
Qwen3.6-35B-A3B-MXFP4_MOE		NVIDIA GeForce RTX 5090	32,768	568.4	573.2	1650.3	51.42	196.69	257	2.22
Qwen3.6-35B-A3B-UD-Q4_K_M	Q4_K_M	NVIDIA GeForce RTX 5090	130,064	558.5	673.1	1541.0	52.12	178.14	299	1.87
Qwen3.5-9B-UD-IQ2_XXS	IQ2_XXS	NVIDIA GeForce RTX 5090	16,384	557.7	424.1	1151.0	54.13	187.56	306	1.82
Qwen3.5-9B-UD-Q3_K_XL	Q3_K_XL	NVIDIA GeForce RTX 5090	16,384	555.9	260.3	994.5	54.69	159.15	321	1.73
Qwen3.5-9B-UD-Q4_K_XL	Q4_K_XL	NVIDIA GeForce RTX 5090	16,384	555.2	195.5	496.2	27.36	109.29	304	1.83
Qwen3.6-35B-A3B-UD-IQ2_M	IQ2_M	NVIDIA GeForce RTX 5090	130,064	554.2	632.9	1584.8	52.72	194.54	297	1.86
Qwen3.5-9B-UD-IQ3_XXS	IQ3_XXS	NVIDIA GeForce RTX 5090	32,768	553.6	303.2	996.5	54.73	162.90	319	1.74
Qwen3.5-9B-IQ4_XS	IQ4_XS	NVIDIA GeForce RTX 5090	32,768	553.5	267.4	1007.8	55.41	190.06	310	1.79
Qwen3.6-35B-A3B-UD-IQ1_M	IQ1_M	NVIDIA GeForce RTX 5090	32,768	553.4	341.5	1089.4	54.20	195.60	265	2.08
Qwen3.6-35B-A3B-UD-IQ2_XXS	IQ2_XXS	NVIDIA GeForce RTX 5090	16,384	552.7	458.5	1447.8	54.24	204.46	299	1.85
Qwen3.5-9B-Q4_0	Q4_0	NVIDIA GeForce RTX 5090	130,064	552.5	251.1	1003.0	55.75	164.88	310	1.78
Qwen3.5-9B-UD-Q5_K_XL	Q5_K_XL	NVIDIA GeForce RTX 5090	130,064	552.1	332.1	999.0	54.65	147.89	321	1.72
Qwen3.5-9B-Q4_1	Q4_1	NVIDIA GeForce RTX 5090	8,192	551.2	286.3	982.0	55.42	183.03	303	1.82
Qwen3.6-35B-A3B-UD-IQ4_NL_XL	IQ4_NL_XL	NVIDIA GeForce RTX 5090	130,064	549.8	445.0	1108.4	53.76	192.66	268	2.05
Qwen3.5-2B-BF16	BF16		130,064	548.6	246.9	636.3	54.08	99.43	—	—
Qwen3.5-9B-UD-IQ2_M	IQ2_M	NVIDIA GeForce RTX 5090	32,768	548.3	665.0	1726.1	54.65	136.66	325	1.69
DeepSeek-R1-Distill-Qwen-32B-Q3_K_M	Q3_K_M	NVIDIA GeForce RTX 5090	32,768	547.8	737.4	1697.4	47.76	187.86	417	1.31
Qwen3.6-35B-A3B-UD-Q2_K_XL	Q2_K_XL	NVIDIA GeForce RTX 5090	130,064	547.4	390.8	1086.0	55.04	171.43	274	2.00
DeepSeek-R1-Distill-Qwen-32B-Q4_K_M	Q4_K_M	NVIDIA GeForce RTX 5090	32,768	547.3	1196.3	3107.8	46.25	181.54	416	1.32
Qwen3.6-35B-A3B-UD-Q3_K_S	Q3_K_S	NVIDIA GeForce RTX 5090	130,064	546.9	429.7	1094.9	54.60	167.03	267	2.05
Qwen3.5-9B-Q3_K_M	Q3_K_M	NVIDIA GeForce RTX 5090	16,384	546.8	262.6	1019.9	55.92	185.02	310	1.76
Qwen3.6-35B-A3B-UD-Q5_K_M	Q5_K_M	NVIDIA GeForce RTX 5090	16,384	546.0	495.1	1539.9	54.55	158.02	273	2.00
Qwen3.6-35B-A3B-UD-Q3_K_XL	Q3_K_XL	NVIDIA GeForce RTX 5090	130,064	545.9	524.0	1397.7	54.10	186.43	294	1.86
Qwen3.6-35B-A3B-UD-Q4_K_XL	Q4_K_XL	NVIDIA GeForce RTX 5090	130,064	545.4	484.8	1234.1	53.99	171.12	268	2.04
Qwen3.5-9B-Q4_K_S	Q4_K_S	NVIDIA GeForce RTX 5090	32,768	544.7	255.4	1019.8	56.34	186.72	312	1.75
Qwen3.5-9B-UD-Q2_K_XL	Q2_K_XL	NVIDIA GeForce RTX 5090	130,064	543.7	499.4	1558.2	55.48	137.49	313	1.74
Qwen3.6-35B-A3B-UD-Q4_K_S	Q4_K_S	NVIDIA GeForce RTX 5090	130,064	543.6	413.7	1107.2	53.84	176.13	266	2.05
Qwen3.6-35B-A3B-UD-Q5_K_S	Q5_K_S	NVIDIA GeForce RTX 5090	32,768	542.1	558.6	1759.7	53.91	165.26	268	2.02
Qwen3.5-9B-Q4_K_M	Q4_K_M	NVIDIA GeForce RTX 5090	130,064	536.3	264.5	1021.8	56.94	160.12	311	1.72
Qwen3.6-35B-A3B-UD-IQ4_NL	IQ4_NL	NVIDIA GeForce RTX 5090	130,064	535.1	571.9	1678.8	54.06	184.48	268	1.99
Qwen3.6-35B-A3B-UD-Q3_K_M	Q3_K_M	NVIDIA GeForce RTX 5090	130,064	534.9	368.1	1091.1	55.31	181.40	264	2.02
Qwen3.6-35B-A3B-UD-IQ4_XS	IQ4_XS	NVIDIA GeForce RTX 5090	130,064	534.7	406.6	1078.9	55.23	142.24	269	1.99
Qwen3.6-35B-A3B-UD-IQ3_XXS	IQ3_XXS	NVIDIA GeForce RTX 5090	16,384	534.3	459.5	1155.7	55.47	144.18	276	1.94
Qwen3.5-9B-Q5_K_M	Q5_K_M	NVIDIA GeForce RTX 5090	16,384	532.5	355.6	1175.2	57.74	185.89	327	1.63
Qwen3.5-9B-Q5_K_S	Q5_K_S	NVIDIA GeForce RTX 5090	16,384	530.9	369.4	1187.8	57.11	201.20	328	1.62
Qwen3.5-2B-UD-IQ3_XXS	IQ3_XXS	NVIDIA GeForce RTX 5060 Ti	130,064	530.7	139.0	357.6	48.58	81.21	99	5.34
Qwen3.5-9B-Q8_0	Q8_0	NVIDIA GeForce RTX 5090	32,768	529.6	256.3	1005.8	58.03	174.28	307	1.73
Qwen3.5-9B-Q6_K	Q6_K	NVIDIA GeForce RTX 5090	16,384	529.1	264.9	1022.2	57.33	183.22	313	1.69
Qwen3.6-35B-A3B-UD-IQ3_S	IQ3_S	NVIDIA GeForce RTX 5090	130,064	526.5	494.1	1459.2	55.92	187.16	277	1.90
Qwen3.5-9B-UD-Q6_K_XL	Q6_K_XL	NVIDIA GeForce RTX 5090	130,064	524.4	280.7	1019.1	57.36	158.28	326	1.61
Qwen3.5-2B-BF16	BF16	NVIDIA GeForce RTX 5090	16,384	518.8	12601.7	14093.3	7.51	33.84	193	2.69
Qwen3.5-9B-UD-Q8_K_XL	Q8_K_XL	NVIDIA GeForce RTX 5090	32,768	513.1	353.8	1176.4	59.20	178.78	338	1.52
Qwen3.5-0.8B-UD-IQ2_XXS	IQ2_XXS	NVIDIA GeForce RTX 4060 Ti	8,192	483.3	217.8	638.9	32.19	34.32	42	11.51
Qwen3.5-9B-BF16	BF16	NVIDIA GeForce RTX 5090	16,384	482.0	376.5	1185.4	62.78	205.51	344	1.40
Qwen3.5-0.8B-Q3_K_M	Q3_K_M	NVIDIA GeForce RTX 4060 Ti	8,192	478.6	268.7	378.8	32.33	35.05	45	10.63
Qwen3.5-0.8B-Q3_K_S	Q3_K_S	NVIDIA GeForce RTX 4060 Ti	16,384	477.2	392.1	806.8	65.37	73.76	43	11.12
Qwen3.5-0.8B-UD-IQ3_XXS	IQ3_XXS	NVIDIA GeForce RTX 4060 Ti	8,192	476.6	258.4	445.1	32.15	36.68	44	10.83
Qwen3.5-0.8B-UD-IQ2_M	IQ2_M	NVIDIA GeForce RTX 4060 Ti	16,384	474.8	326.8	670.1	64.53	75.41	43	11.04
Qwen3.5-0.8B-UD-Q2_K_XL	Q2_K_XL	NVIDIA GeForce RTX 4060 Ti	16,384	473.4	352.6	783.7	65.16	79.18	44	10.88
Qwen3.5-0.8B-IQ4_XS	IQ4_XS	NVIDIA GeForce RTX 4060 Ti	16,384	471.9	302.9	687.4	65.10	88.93	43	10.95
Qwen3.5-0.8B-Q4_0	Q4_0	NVIDIA GeForce RTX 4060 Ti	16,384	471.4	269.0	609.0	65.05	88.67	44	10.84
Qwen3.5-0.8B-Q4_1	Q4_1	NVIDIA GeForce RTX 4060 Ti	16,384	470.5	302.1	747.7	65.61	85.66	44	10.82
Qwen3.5-0.8B-Q4_K_S	Q4_K_S	NVIDIA GeForce RTX 4060 Ti	16,384	469.3	263.2	619.9	65.53	90.25	44	10.74
Qwen3.5-0.8B-IQ4_NL	IQ4_NL	NVIDIA GeForce RTX 4060 Ti	16,384	467.4	256.9	620.5	65.53	88.91	43	10.82
Qwen3.5-0.8B-UD-Q5_K_XL	Q5_K_XL	NVIDIA GeForce RTX 4060 Ti	16,384	466.9	239.1	619.5	66.21	91.87	44	10.64
Qwen3.5-0.8B-Q4_K_M	Q4_K_M	NVIDIA GeForce RTX 4060 Ti	16,384	466.7	232.8	555.7	65.73	94.04	44	10.73
Qwen3.5-0.8B-UD-Q4_K_XL	Q4_K_XL	NVIDIA GeForce RTX 4060 Ti	32,768	464.8	338.6	733.8	65.76	88.91	44	10.64
Qwen3.5-2B-UD-IQ2_M	IQ2_M	NVIDIA GeForce RTX 5060 Ti	32,768	464.7	140.3	306.7	52.76	96.32	82	5.65
Qwen3.5-0.8B-UD-Q3_K_XL	Q3_K_XL	NVIDIA GeForce RTX 4060 Ti	16,384	463.6	252.0	645.2	66.02	89.27	45	10.35
Qwen3.5-0.8B-Q5_K_M	Q5_K_M	NVIDIA GeForce RTX 4060 Ti	16,384	462.8	263.1	632.5	66.11	89.85	45	10.40
Qwen3.5-0.8B-Q5_K_S	Q5_K_S	NVIDIA GeForce RTX 4060 Ti	32,768	462.5	261.8	633.1	66.38	89.96	44	10.49
Qwen3.5-0.8B-Q6_K	Q6_K	NVIDIA GeForce RTX 4060 Ti	16,384	460.7	233.9	542.6	65.90	90.54	44	10.42
DeepSeek-R1-Distill-Qwen-32B-Q5_K_M	Q5_K_M	NVIDIA GeForce RTX 5090	16,384	459.3	858.3	1901.6	53.84	203.96	433	1.06
Qwen3.5-0.8B-Q8_0	Q8_0	NVIDIA GeForce RTX 4060 Ti	16,384	457.8	252.5	615.1	66.88	94.62	45	10.17
Qwen3.6-35B-A3B-UD-Q5_K_XL	Q5_K_XL	NVIDIA GeForce RTX 5090	16,384	457.6	428.9	1098.5	64.00	216.36	252	1.82
Qwen3.5-0.8B-UD-Q6_K_XL	Q6_K_XL	NVIDIA GeForce RTX 4060 Ti	16,384	451.7	266.0	615.0	67.73	91.78	45	10.04
Qwen3.5-2B-UD-IQ2_XXS	IQ2_XXS	NVIDIA GeForce RTX 4060 Ti	32,768	450.9	470.6	971.7	69.35	73.53	45	10.02
Qwen3.5-4B-Q4_0	Q4_0	NVIDIA GeForce RTX 5090	32,768	446.6	2260.6	2525.2	8.35	10.88	—	—
Qwen3.5-2B-UD-IQ2_M	IQ2_M	NVIDIA GeForce RTX 4060 Ti	32,768	446.5	465.8	954.3	70.00	74.78	46	9.73
Qwen3.5-2B-Q3_K_M	Q3_K_M	NVIDIA GeForce RTX 4060 Ti	32,768	438.7	366.1	845.0	70.53	92.93	47	9.35
Qwen3.5-4B-Q4_1	Q4_1	NVIDIA GeForce RTX 5090	8,192	437.7	2208.7	2511.0	8.76	29.14	—	—
Qwen3.5-2B-UD-Q2_K_XL	Q2_K_XL	NVIDIA GeForce RTX 4060 Ti	16,384	437.6	448.7	972.2	70.32	85.67	47	9.25
Qwen3.5-0.8B-UD-Q8_K_XL	Q8_K_XL	NVIDIA GeForce RTX 4060 Ti	32,768	435.3	254.1	504.8	70.05	99.03	46	9.50
Qwen3.5-4B-IQ4_NL	IQ4_NL	NVIDIA GeForce RTX 5090	8,192	434.5	2236.7	2490.2	8.78	58.74	—	—
Qwen3.5-4B-IQ4_XS	IQ4_XS	NVIDIA GeForce RTX 5090	3,055	433.1	200.8	294.3	8.45	10.96	253	1.71
Qwen3.5-2B-Q4_K_M	Q4_K_M	NVIDIA GeForce RTX 4060 Ti	16,384	432.9	285.0	804.8	71.36	96.11	47	9.15
Qwen3.5-2B-UD-IQ3_XXS	IQ3_XXS	NVIDIA GeForce RTX 4060 Ti	32,768	432.5	378.0	878.3	71.72	93.16	48	9.11
Qwen3.5-2B-IQ4_XS	IQ4_XS	NVIDIA GeForce RTX 4060 Ti	32,768	431.9	324.7	844.1	71.24	102.37	47	9.19
Qwen3.5-2B-Q4_0	Q4_0	NVIDIA GeForce RTX 4060 Ti	16,384	431.1	307.5	879.1	71.31	99.36	47	9.15
Qwen3.5-2B-Q4_K_S	Q4_K_S	NVIDIA GeForce RTX 4060 Ti	130,064	430.9	355.8	783.0	71.41	101.27	47	9.17
Qwen3.5-2B-IQ4_NL	IQ4_NL	NVIDIA GeForce RTX 4060 Ti	16,384	430.3	277.4	805.9	71.70	109.25	46	9.36
Qwen3.5-2B-UD-Q3_K_XL	Q3_K_XL	NVIDIA GeForce RTX 4060 Ti	32,768	429.4	317.4	827.7	72.19	100.46	48	9.00
Qwen3.5-4B-UD-IQ2_XXS	IQ2_XXS	NVIDIA GeForce RTX 5090	8,192	428.7	2472.8	2676.4	8.57	11.17	—	—
Qwen3.5-2B-Q4_1	Q4_1	NVIDIA GeForce RTX 4060 Ti	16,384	428.5	301.9	897.5	71.81	100.85	45	9.48
Qwen3.5-2B-Q3_K_S	Q3_K_S	NVIDIA GeForce RTX 4060 Ti	8,192	428.1	270.3	343.6	35.98	43.20	47	9.13
Qwen3.5-4B-UD-Q2_K_XL	Q2_K_XL	NVIDIA GeForce RTX 5090	16,384	427.3	2392.6	2636.8	8.56	13.88	—	—
Qwen3.5-0.8B-BF16	BF16	NVIDIA GeForce RTX 4060 Ti	16,384	425.1	312.8	905.4	71.58	111.96	42	10.10
Qwen3.5-2B-UD-Q4_K_XL	Q4_K_XL	NVIDIA GeForce RTX 4060 Ti	32,768	424.8	300.9	739.1	72.20	111.07	48	8.92
Qwen3.5-2B-UD-Q5_K_XL	Q5_K_XL	NVIDIA GeForce RTX 4060 Ti	16,384	422.2	310.2	946.6	72.92	106.31	48	8.87
Qwen3.5-2B-Q5_K_S	Q5_K_S	NVIDIA GeForce RTX 4060 Ti	16,384	421.3	294.8	862.7	72.83	106.24	48	8.70
Qwen3.5-2B-Q5_K_M	Q5_K_M	NVIDIA GeForce RTX 4060 Ti	16,384	419.0	344.4	967.4	73.34	100.59	49	8.52
Qwen3.5-4B-Q3_K_S	Q3_K_S	NVIDIA GeForce RTX 5090	8,192	417.4	207.2	305.4	8.79	11.23	—	—
Qwen3.5-2B-Q6_K	Q6_K	NVIDIA GeForce RTX 4060 Ti	16,384	417.3	330.4	883.2	73.78	101.97	49	8.60
Qwen3.5-35B-A3B-UD-Q4_K_L	Q4_K_L	NVIDIA GeForce RTX 5090	32,768	415.6	575.1	1559.6	72.18	214.21	338	1.23
Qwen3.5-4B-UD-IQ2_M	IQ2_M	NVIDIA GeForce RTX 5090	130,064	407.6	2344.9	2698.7	9.36	11.63	—	—
Qwen3.5-2B-Q8_0	Q8_0	NVIDIA GeForce RTX 4060 Ti	130,064	406.8	328.2	955.2	75.61	115.53	46	8.94
Qwen3.5-35B-A3B-MXFP4_MOE		NVIDIA GeForce RTX 5090	16,384	405.6	518.6	1478.6	74.06	209.11	337	1.20
Qwen3.5-4B-Q6_K	Q6_K	NVIDIA GeForce RTX 5090	130,064	403.6	2348.2	2701.7	9.51	14.26	—	—
Qwen3.5-35B-A3B-APEX-I-Compact		NVIDIA GeForce RTX 5090	32,768	402.9	565.3	1617.9	75.05	242.18	335	1.20
Qwen3.5-35B-A3B-APEX-Quality	Quality	NVIDIA GeForce RTX 5090	130,064	401.9	505.8	1504.5	75.37	179.71	336	1.20
Qwen3.5-2B-UD-Q6_K_XL	Q6_K_XL	NVIDIA GeForce RTX 4060 Ti	16,384	401.1	304.5	756.8	76.61	120.74	50	8.10
Qwen3.5-35B-A3B-APEX-I-Quality	Quality	NVIDIA GeForce RTX 5090	130,064	400.8	519.2	1624.8	75.42	214.32	336	1.19
Qwen3.5-4B-UD-IQ3_XXS	IQ3_XXS	NVIDIA GeForce RTX 5090	32,768	400.4	2396.9	2733.1	9.59	41.83	—	—
Qwen3.5-35B-A3B-UD-IQ2_M	IQ2_M	NVIDIA GeForce RTX 5090	16,384	400.4	523.1	1373.6	75.70	248.97	354	1.13
Qwen3.5-35B-A3B-Q4_K_M	Q4_K_M	NVIDIA GeForce RTX 5090	130,064	400.1	561.6	1963.8	75.46	208.36	326	1.23
Qwen3.5-35B-A3B-UD-Q4_K_XL	Q4_K_XL	NVIDIA GeForce RTX 5090	32,768	399.8	570.2	1620.8	75.34	190.59	326	1.23
Qwen3.5-35B-A3B-UD-Q3_K_XL	Q3_K_XL	NVIDIA GeForce RTX 5090	32,768	399.7	537.3	1554.2	75.70	225.77	327	1.22
Qwen3.5-35B-A3B-APEX-Compact		NVIDIA GeForce RTX 5090	130,064	399.1	483.6	1585.2	75.84	186.23	333	1.20
Qwen3.5-35B-A3B-Q3_K_M	Q3_K_M	NVIDIA GeForce RTX 5090	16,384	398.1	583.0	2062.3	75.91	179.72	347	1.15
Qwen3.5-2B-UD-Q2_K_XL	Q2_K_XL	NVIDIA GeForce RTX 5060 Ti	130,064	397.6	147.6	269.0	59.47	108.67	78	5.11
Qwen3.5-35B-A3B-Q4_K_S	Q4_K_S	NVIDIA GeForce RTX 5090	16,384	397.6	427.3	1291.2	76.04	238.72	345	1.15
Qwen3.5-35B-A3B-Q5_K_S	Q5_K_S	NVIDIA GeForce RTX 5090	32,768	397.1	425.7	1315.5	76.36	245.60	328	1.21
Qwen3.5-0.8B-Q4_0	Q4_0	NVIDIA GB10	2,048	395.6	123.7	236.3	9.57	12.14	—	—
Qwen3.5-35B-A3B-Q3_K_S	Q3_K_S	NVIDIA GeForce RTX 5090	130,064	395.2	457.1	1482.0	76.99	190.91	326	1.21
Qwen3.5-35B-A3B-UD-IQ2_XXS	IQ2_XXS	NVIDIA GeForce RTX 5090	16,384	394.6	370.2	1123.6	77.30	214.38	350	1.13
Qwen3.5-4B-Q8_0	Q8_0	NVIDIA GeForce RTX 5090	32,768	393.8	7314.2	8046.0	9.80	18.15	—	—
Qwen3.5-0.8B-UD-IQ2_XXS	IQ2_XXS	NVIDIA GB10	32,768	393.3	135.2	234.3	9.67	12.44	—	—
Qwen3.5-35B-A3B-UD-IQ4_XS	IQ4_XS	NVIDIA GeForce RTX 5090	130,064	393.0	541.4	1675.2	76.86	202.32	326	1.21
Qwen3.5-0.8B-Q4_1	Q4_1	NVIDIA GB10	2,048	393.0	119.9	237.7	9.62	12.26	—	—
Qwen3.5-35B-A3B-UD-Q2_K_XL	Q2_K_XL	NVIDIA GeForce RTX 5090	32,768	391.6	592.8	2003.4	77.00	189.95	330	1.19
Qwen3.5-35B-A3B-UD-IQ4_NL	IQ4_NL	NVIDIA GeForce RTX 5090	32,768	390.9	508.8	1556.5	77.14	225.04	327	1.20
Qwen3.5-35B-A3B-UD-IQ3_XXS	IQ3_XXS	NVIDIA GeForce RTX 5090	32,768	390.6	702.0	2320.4	77.19	191.16	331	1.18
Qwen3.5-35B-A3B-UD-IQ3_S	IQ3_S	NVIDIA GeForce RTX 5090	32,768	390.5	538.2	1652.8	77.68	190.04	332	1.18
Qwen3.5-35B-A3B-APEX-I-Balanced	Balanced	NVIDIA GeForce RTX 5090	16,384	390.2	417.8	1250.9	77.59	231.99	335	1.16
Qwen3.5-35B-A3B-APEX-Mini		NVIDIA GeForce RTX 5090	130,064	389.5	392.2	1162.7	78.40	223.97	333	1.17
Qwen3.5-35B-A3B-APEX-Balanced	Balanced	NVIDIA GeForce RTX 5090	32,768	389.0	653.3	1597.2	77.13	213.84	362	1.08
Qwen3.5-0.8B-UD-Q2_K_XL	Q2_K_XL	NVIDIA GB10	65,536	386.5	137.7	239.6	9.84	12.74	—	—
Qwen3.5-0.8B-IQ4_NL	IQ4_NL	NVIDIA GB10	32,768	386.2	110.0	194.9	9.65	15.20	—	—
Qwen3.5-4B-Q3_K_M	Q3_K_M	NVIDIA GeForce RTX 5090	32,768	386.1	16752.7	18880.4	10.07	59.59	—	—
Qwen3.5-0.8B-Q3_K_S	Q3_K_S	NVIDIA GB10	8,192	385.0	133.8	235.5	9.90	12.43	—	—
Qwen3.5-4B-UD-Q3_K_XL	Q3_K_XL	NVIDIA GeForce RTX 5090	16,384	384.3	7361.5	8196.9	10.07	48.91	—	—
Qwen3.5-0.8B-UD-IQ2_M	IQ2_M	NVIDIA GB10	8,192	381.6	137.2	238.0	9.98	12.83	—	—
Qwen3.5-9B-IQ4_NL	IQ4_NL	NVIDIA GeForce RTX 5090	130,064	380.4	2512.7	2851.5	10.15	56.49	218	1.75
Qwen3.5-4B-IQ4_XS	IQ4_XS	NVIDIA GeForce RTX 5060 Ti	32,768	379.4	584.9	1920.4	79.46	172.02	97	3.92
Qwen3.5-4B-Q4_0	Q4_0	NVIDIA GeForce RTX 5060 Ti	16,384	378.9	530.4	1473.1	79.97	186.95	103	3.69
Qwen3.5-4B-IQ4_NL	IQ4_NL	NVIDIA GeForce RTX 5060 Ti	130,064	377.1	517.5	1669.7	79.75	180.16	97	3.91
Qwen3.5-4B-Q3_K_S	Q3_K_S	NVIDIA GeForce RTX 5060 Ti	16,384	376.8	664.1	2165.2	80.52	178.56	103	3.64
Qwen3.5-2B-UD-Q8_K_XL	Q8_K_XL	NVIDIA GeForce RTX 4060 Ti	16,384	376.7	351.5	915.7	81.13	122.87	47	8.07
Qwen3.5-0.8B-IQ4_XS	IQ4_XS	NVIDIA GB10	65,536	376.1	126.9	227.7	10.04	12.76	—	—
Qwen3.5-0.8B-UD-IQ3_XXS	IQ3_XXS	NVIDIA GB10	2,048	374.5	127.5	233.1	10.21	13.06	—	—
Qwen3.5-4B-UD-Q6_K_XL	Q6_K_XL	NVIDIA GeForce RTX 5090	32,768	373.0	17685.5	19603.0	10.39	53.32	267	1.40
Qwen3.5-4B-Q4_1	Q4_1	NVIDIA GeForce RTX 5060 Ti	32,768	372.6	521.2	1756.3	80.69	171.40	96	3.87
Qwen3.5-35B-A3B-Q5_K_M	Q5_K_M	NVIDIA GeForce RTX 5090	16,384	371.3	685.2	2126.1	81.15	201.86	332	1.12
Qwen3.5-4B-Q3_K_M	Q3_K_M	NVIDIA GeForce RTX 5060 Ti	32,768	371.1	646.1	1894.1	80.55	190.10	104	3.56
Qwen3.5-4B-Q4_K_M	Q4_K_M	NVIDIA GeForce RTX 5060 Ti	32,768	368.7	607.1	2012.8	81.30	185.04	98	3.75
Qwen3.5-0.8B-Q3_K_M	Q3_K_M	NVIDIA GB10	65,536	368.4	112.0	201.7	10.19	14.66	—	—
Qwen3.5-0.8B-Q4_K_S	Q4_K_S	NVIDIA GB10	32,768	364.8	2583.8	2964.5	10.63	44.87	—	—
Qwen3.5-0.8B-Q4_K_M	Q4_K_M	NVIDIA GB10	130,064	364.1	2640.7	2986.4	10.57	13.57	—	—
Qwen3.5-35B-A3B-UD-Q5_K_XL	Q5_K_XL	NVIDIA GeForce RTX 5090	8,192	363.8	331.7	919.9	41.70	120.87	323	1.13
Qwen3.5-4B-UD-Q5_K_XL	Q5_K_XL	NVIDIA GeForce RTX 5090	8,192	363.1	7839.8	8633.4	10.61	43.61	—	—
Qwen3.5-4B-Q4_K_M	Q4_K_M	NVIDIA GeForce RTX 5090	32,768	360.2	2755.1	3037.4	10.62	12.97	—	—
DeepSeek-R1-Distill-Qwen-32B-Q6_K	Q6_K	NVIDIA GeForce RTX 5090	8,192	359.4	385.3	1001.3	38.06	138.92	436	0.83
Qwen3.5-0.8B-Q5_K_M	Q5_K_M	NVIDIA GB10	32,768	358.6	2666.2	3037.4	10.78	13.85	—	—
Qwen3.5-4B-Q5_K_M	Q5_K_M	NVIDIA GeForce RTX 5090	16,384	358.4	138.3	218.7	10.55	12.84	—	—
Qwen3.5-2B-BF16	BF16	NVIDIA GeForce RTX 4060 Ti	16,384	357.6	382.6	1137.6	85.23	135.69	50	7.22
Qwen3.5-0.8B-Q5_K_S	Q5_K_S	NVIDIA GB10	2,048	356.6	2612.0	3052.6	10.90	14.67	—	—
Qwen3.5-4B-UD-Q4_K_XL	Q4_K_XL	NVIDIA GeForce RTX 5090	8,192	356.3	2685.1	3053.3	10.79	44.57	—	—
Qwen3.5-0.8B-UD-Q4_K_XL	Q4_K_XL	NVIDIA GB10	65,536	355.5	111.5	194.9	10.63	13.73	—	—
Qwen3.5-0.8B-UD-Q3_K_XL	Q3_K_XL	NVIDIA GB10	130,064	355.1	2451.6	3005.3	10.87	47.80	—	—
Qwen3.5-4B-Q4_K_S	Q4_K_S	NVIDIA GeForce RTX 5090	32,768	354.3	7983.5	8846.7	10.90	49.01	—	—
Qwen3.5-0.8B-UD-Q5_K_XL	Q5_K_XL	NVIDIA GB10	2,048	350.7	2667.5	3086.1	11.09	14.48	—	—
gemma-4-E2B-it-IQ4_NL	IQ4_NL	Apple M4 Pro	32,768	349.7	393.2	940.8	80.29	182.73	28	12.58
Qwen3.6-27B-IQ4_XS	IQ4_XS	NVIDIA GeForce RTX 5090	16,384	349.1	1110.7	3378.9	83.69	345.09	380	0.92
Qwen3.5-4B-UD-Q8_K_XL	Q8_K_XL	NVIDIA GeForce RTX 5090	32,768	349.1	8065.5	9006.7	11.11	52.61	209	1.67
Qwen3.5-4B-Q5_K_S	Q5_K_S	NVIDIA GeForce RTX 5090	8,192	347.9	18868.0	20994.9	11.15	49.70	—	—
gemma-4-26B-A4B-it-UD-IQ2_M	IQ2_M	NVIDIA GB10	65,536	341.9	1306.0	4706.0	86.83	185.16	49	6.94
gemma-4-E2B-it-UD-Q4_K_XL	Q4_K_XL	Apple M4 Pro	16,384	341.1	465.8	968.6	84.53	243.70	28	12.10
Qwen3.5-0.8B-Q6_K	Q6_K	NVIDIA GB10	130,064	341.1	93.5	178.1	11.08	17.92	—	—
gemma-4-E2B-it-Q8_0	Q8_0	Apple M4 Pro	32,768	339.4	401.3	955.1	78.42	187.70	30	11.16
gemma-4-E2B-it-UD-Q8_K_XL	Q8_K_XL	Apple M4 Pro	130,064	338.2	399.3	983.2	79.88	198.39	30	11.24
Qwen3.5-0.8B-UD-Q6_K_XL	Q6_K_XL	NVIDIA GB10	130,064	337.1	2799.9	3148.4	11.48	17.62	—	—
gemma-4-26B-A4B-it-UD-IQ2_XXS	IQ2_XXS	NVIDIA GB10	32,768	336.8	1001.3	3842.2	89.35	226.12	46	7.26
Qwen3.5-0.8B-Q8_0	Q8_0	NVIDIA GB10	130,064	335.4	2761.9	3214.3	11.59	47.50	—	—
Qwen3.6-35B-A3B-UD-Q6_K	Q6_K	NVIDIA GeForce RTX 5090	8,192	333.4	343.9	708.9	44.63	140.03	233	1.43
gemma-4-E2B-it-IQ4_XS	IQ4_XS	Apple M4 Pro	130,064	333.3	418.6	995.4	81.38	194.05	30	11.30
Qwen3.5-9B-Q3_K_S	Q3_K_S	NVIDIA GeForce RTX 5090	32,768	332.6	2881.3	3280.6	11.65	62.40	216	1.54
gemma-4-E2B-it-UD-Q3_K_XL	Q3_K_XL	Apple M4 Pro	16,384	330.4	424.1	935.5	82.83	213.39	28	11.72
Qwen3.5-9B-IQ4_XS	IQ4_XS	NVIDIA GeForce RTX 5060 Ti	130,064	329.0	459.8	1001.9	92.77	253.36	109	3.01
gemma-4-E2B-it-Q4_1	Q4_1	Apple M4 Pro	130,064	328.8	469.4	1037.4	82.67	227.10	33	9.99
Qwen3.5-9B-IQ4_NL	IQ4_NL	NVIDIA GeForce RTX 5060 Ti	16,384	328.3	498.8	1255.2	93.56	248.88	116	2.82
Qwen3.5-9B-Q4_0	Q4_0	NVIDIA GeForce RTX 5060 Ti	130,064	328.1	411.5	1016.5	93.65	215.59	109	3.02
Qwen3.6-27B-Q5_K_S	Q5_K_S	NVIDIA GeForce RTX 5090	32,768	327.6	971.5	2643.1	89.78	454.18	387	0.85
gemma-4-E2B-it-Q4_0	Q4_0	Apple M4 Pro	32,768	327.6	444.1	900.8	83.03	288.46	30	10.92
gemma-4-E2B-it-UD-Q6_K_XL	Q6_K_XL	Apple M4 Pro	32,768	327.3	434.2	1006.2	83.14	212.35	30	10.98
Qwen3.5-27B-IQ4_XS	IQ4_XS	NVIDIA GeForce RTX 5090	32,768	326.2	906.2	2616.6	90.92	451.66	289	1.13
Qwen3.5-9B-UD-IQ2_XXS	IQ2_XXS	NVIDIA GeForce RTX 5060 Ti	32,768	326.0	1019.1	3721.3	90.12	190.94	107	3.05
Qwen3.5-2B-UD-Q3_K_XL	Q3_K_XL	NVIDIA GeForce RTX 5060 Ti	32,768	325.7	157.3	259.5	59.87	118.07	90	3.61
Qwen3.6-27B-Q4_K_M	Q4_K_M	NVIDIA GeForce RTX 5090	32,768	321.4	957.6	2498.6	91.04	465.86	316	1.02
Qwen3.5-9B-Q4_1	Q4_1	NVIDIA GeForce RTX 5060 Ti	32,768	321.3	561.6	1325.9	94.27	231.01	115	2.79
gemma-4-E2B-it-Q6_K	Q6_K	Apple M4 Pro	16,384	320.9	454.1	1021.2	86.01	236.56	29	11.07
gemma-4-E2B-it-UD-Q2_K_XL	Q2_K_XL	Apple M4 Pro	130,064	320.3	442.1	1012.0	84.11	210.51	28	11.56
Qwen3.5-2B-UD-IQ2_XXS	IQ2_XXS	NVIDIA GB10	130,064	320.1	3174.6	3538.0	12.02	14.87	62	5.15
Qwen3.5-9B-Q4_K_S	Q4_K_S	NVIDIA GeForce RTX 5060 Ti	32,768	319.4	536.9	1633.0	94.73	239.94	110	2.91
Qwen3.5-9B-UD-Q4_K_XL	Q4_K_XL	NVIDIA GeForce RTX 5060 Ti	130,064	319.4	562.1	1978.5	94.75	232.04	111	2.89
gemma-4-E2B-it-Q5_K_S	Q5_K_S	Apple M4 Pro	130,064	318.9	495.2	1079.3	86.59	265.72	29	11.07
Qwen3.5-9B-Q4_K_M	Q4_K_M	NVIDIA GeForce RTX 5060 Ti	130,064	318.7	564.4	1296.1	95.71	274.25	119	2.67
Qwen3.5-27B-IQ4_NL	IQ4_NL	NVIDIA GeForce RTX 5090	32,768	318.4	1304.6	4160.4	90.78	338.05	291	1.09
Qwen3.5-35B-A3B-UD-Q6_K_S	Q6_K_S	NVIDIA GeForce RTX 5090	8,192	318.1	408.0	1036.0	47.26	147.07	324	0.98
Qwen3.5-9B-Q3_K_M	Q3_K_M	NVIDIA GeForce RTX 5060 Ti	16,384	317.7	435.6	972.5	96.87	236.44	110	2.90
gemma-4-E2B-it-Q4_K_M	Q4_K_M	Apple M4 Pro	130,064	314.9	477.5	1093.5	87.83	242.40	33	9.57
gemma-4-E2B-it-UD-Q5_K_XL	Q5_K_XL	Apple M4 Pro	130,064	314.8	448.4	1031.1	86.70	216.51	29	10.89
Qwen3.5-9B-Q3_K_S	Q3_K_S	NVIDIA GeForce RTX 5060 Ti	32,768	314.4	547.3	1349.3	96.46	251.58	117	2.68
Qwen3.5-4B-BF16	BF16	NVIDIA GeForce RTX 5090	130,064	313.2	8847.9	9966.8	12.37	62.68	—	—
gemma-4-E2B-it-Q3_K_S	Q3_K_S	Apple M4 Pro	130,064	313.1	479.4	1096.5	88.77	243.28	30	10.30
gemma-4-E2B-it-BF16	BF16	Apple M4 Pro	130,064	312.0	424.2	949.9	84.78	213.44	32	9.66
Qwen3.6-27B-Q4_K_S	Q4_K_S	NVIDIA GeForce RTX 5090	32,768	311.6	1403.6	3455.7	91.36	429.63	333	0.94
Qwen3.5-27B-Q3_K_S	Q3_K_S	NVIDIA GeForce RTX 5090	32,768	311.5	875.6	2650.3	94.89	440.04	295	1.06
gemma-4-E2B-it-Q3_K_M	Q3_K_M	Apple M4 Pro	32,768	311.2	458.1	1020.8	87.78	229.36	29	10.73
Qwen3.5-9B-UD-Q5_K_XL	Q5_K_XL	NVIDIA GeForce RTX 5060 Ti	130,064	311.1	485.5	999.4	97.97	272.59	113	2.76
Qwen3.5-2B-UD-IQ2_XXS	IQ2_XXS	Apple M4 Pro	8,192	309.0	597.2	1535.5	101.21	121.51	28	11.08
Qwen3.5-2B-UD-IQ2_M	IQ2_M	NVIDIA GB10	8,192	308.7	144.9	284.8	12.43	15.22	62	5.01
Qwen3.5-27B-Q3_K_M	Q3_K_M	NVIDIA GeForce RTX 5090	16,384	307.4	1013.5	2720.9	94.91	383.23	294	1.05
Qwen3.5-9B-Q5_K_S	Q5_K_S	NVIDIA GeForce RTX 5060 Ti	130,064	306.5	479.7	1027.5	99.61	278.49	115	2.67
Qwen3.6-27B-Q3_K_S	Q3_K_S	NVIDIA GeForce RTX 5090	130,064	306.4	965.3	2518.7	96.51	452.15	323	0.95
Qwen3.5-35B-A3B-Q6_K	Q6_K	NVIDIA GeForce RTX 5090	8,192	303.0	310.1	689.4	49.85	172.74	310	0.98
gemma-4-E2B-it-Q4_K_S	Q4_K_S	Apple M4 Pro	16,384	302.6	446.6	1089.5	89.34	218.73	32	9.46
gemma-4-E2B-it-Q5_K_M	Q5_K_M	Apple M4 Pro	130,064	302.5	497.4	946.9	90.20	338.92	28	10.96
Qwen3.5-9B-Q6_K	Q6_K	NVIDIA GeForce RTX 5060 Ti	32,768	302.0	544.9	1358.8	100.60	266.50	122	2.48
gemma-4-E2B-it-UD-IQ2_M	IQ2_M	Apple M4 Pro	130,064	301.7	433.3	1023.2	87.74	202.45	29	10.26
Qwen3.5-9B-Q5_K_M	Q5_K_M	NVIDIA GeForce RTX 5060 Ti	32,768	300.9	479.7	1110.5	100.29	247.31	118	2.56
Qwen3.5-27B-Q4_0	Q4_0	NVIDIA GeForce RTX 5090	32,768	300.2	970.9	2738.6	98.67	490.11	278	1.08
Qwen3.6-27B-Q5_K_M	Q5_K_M	NVIDIA GeForce RTX 5090	16,384	299.9	934.0	2530.4	97.40	434.12	386	0.78
Qwen3.5-27B-Q4_1	Q4_1	NVIDIA GeForce RTX 5090	32,768	297.3	1498.3	4344.9	97.56	337.04	278	1.07
Qwen3.6-27B-Q2_K	Q2_K	NVIDIA GeForce RTX 5090	130,064	297.0	887.9	3341.8	94.69	351.55	326	0.91
Qwen3.5-0.8B-UD-Q8_K_XL	Q8_K_XL	NVIDIA GB10	130,064	296.9	3119.1	3564.2	13.07	49.17	—	—
Qwen3.6-27B-Q3_K_M	Q3_K_M	NVIDIA GeForce RTX 5090	16,384	296.5	848.3	2508.7	99.37	365.19	322	0.92
Qwen3.5-2B-UD-IQ2_M	IQ2_M	Apple M4 Pro	8,192	295.7	646.2	1479.2	105.58	135.14	28	10.41
Qwen3.5-2B-UD-IQ3_XXS	IQ3_XXS	NVIDIA GB10	65,536	295.6	141.5	273.5	12.96	15.82	60	4.93
Qwen3.5-9B-Q8_0	Q8_0	NVIDIA GeForce RTX 5060 Ti	32,768	294.7	525.0	1282.3	103.56	256.47	114	2.59
Qwen3.5-2B-UD-Q2_K_XL	Q2_K_XL	NVIDIA GB10	130,064	294.6	146.4	289.7	13.05	15.88	60	4.95
gemma-4-E2B-it-UD-IQ3_XXS	IQ3_XXS	Apple M4 Pro	16,384	294.3	423.8	975.9	88.55	214.32	30	9.84
Qwen3.5-2B-Q4_0	Q4_0	NVIDIA GB10	130,064	293.1	144.4	273.4	13.13	16.00	53	5.50
Qwen3.5-27B-Q4_K_S	Q4_K_S	NVIDIA GeForce RTX 5090	32,768	290.4	1096.8	3548.7	101.05	422.53	285	1.02
Qwen3.5-27B-Q4_K_M	Q4_K_M	NVIDIA GeForce RTX 5090	16,384	288.5	1570.4	4749.0	99.98	387.41	288	1.00
Qwen3.6-27B-Q3_K_L	Q3_K_L	NVIDIA GeForce RTX 5090	32,768	286.9	879.4	2487.7	101.32	394.56	326	0.88
Qwen3.5-2B-Q3_K_S	Q3_K_S	NVIDIA GB10	2,048	283.6	3472.8	3868.9	13.56	16.52	63	4.50
Qwen3.5-2B-Q3_K_M	Q3_K_M	NVIDIA GB10	130,064	283.5	143.3	283.4	13.60	16.59	60	4.69
Qwen3.5-2B-IQ4_NL	IQ4_NL	NVIDIA GB10	2,048	283.4	3319.1	3787.1	13.75	49.96	57	5.00
Qwen3.5-2B-Q3_K_S	Q3_K_S	Apple M4 Pro	32,768	283.3	750.9	2701.4	108.55	185.93	28	9.97
Qwen3.5-2B-UD-IQ3_XXS	IQ3_XXS	Apple M4 Pro	32,768	282.7	735.2	3176.0	108.56	170.36	30	9.52
Qwen3.5-2B-Q3_K_M	Q3_K_M	Apple M4 Pro	8,192	282.3	637.2	1456.2	108.09	181.40	29	9.80
Qwen3.5-4B-BF16	BF16	NVIDIA GeForce RTX 5060 Ti	32,768	281.6	728.4	2537.1	103.87	259.15	101	2.80
Qwen3.5-2B-Q4_1	Q4_1	NVIDIA GB10	2,048	278.9	3377.3	3809.6	13.89	48.13	55	5.04
Qwen3.5-2B-IQ4_NL	IQ4_NL	Apple M4 Pro	130,064	277.7	517.5	2378.3	110.43	218.43	29	9.51
Qwen3.5-2B-Q4_0	Q4_0	Apple M4 Pro	8,192	277.4	557.4	1308.9	109.27	184.02	29	9.56
Qwen3.5-2B-Q8_0	Q8_0	Apple M4 Pro	8,192	276.3	457.8	833.3	110.18	203.17	30	9.30
Qwen3.5-4B-UD-IQ2_XXS	IQ2_XXS	NVIDIA GeForce RTX 4060 Ti	32,768	275.9	982.0	2899.3	110.56	173.71	52	5.27
gemma-4-26B-A4B-it-UD-IQ3_S	IQ3_S	NVIDIA GB10	32,768	274.7	835.1	3381.8	106.45	291.62	48	5.71
Qwen3.5-2B-UD-Q3_K_XL	Q3_K_XL	NVIDIA GB10	65,536	274.3	146.4	278.4	14.06	17.15	58	4.74
Qwen3.5-2B-UD-Q3_K_XL	Q3_K_XL	Apple M4 Pro	8,192	273.8	766.9	1468.7	110.22	197.07	29	9.34
Qwen3.5-2B-IQ4_XS	IQ4_XS	NVIDIA GB10	8,192	273.7	3464.7	3917.6	14.15	17.61	55	5.00
Qwen3.5-27B-Q5_K_S	Q5_K_S	NVIDIA GeForce RTX 5090	16,384	273.4	1364.7	3654.7	106.41	510.92	309	0.88
Qwen3.5-2B-Q4_1	Q4_1	Apple M4 Pro	8,192	273.3	648.7	1511.4	110.34	190.13	29	9.39
Qwen3.5-2B-Q4_K_S	Q4_K_S	NVIDIA GB10	8,192	273.0	114.7	230.1	13.95	20.31	61	4.48
Qwen3.5-2B-UD-Q2_K_XL	Q2_K_XL	Apple M4 Pro	16,384	272.5	896.0	4434.5	111.37	184.21	27	10.13
Qwen3.5-2B-UD-Q6_K_XL	Q6_K_XL	Apple M4 Pro	32,768	271.4	551.7	2447.2	112.57	214.84	30	9.11
Qwen3.5-2B-UD-Q8_K_XL	Q8_K_XL	Apple M4 Pro	32,768	270.7	642.8	3033.8	111.04	185.95	31	8.85
Qwen3.6-35B-A3B-UD-Q6_K_XL	Q6_K_XL	NVIDIA GeForce RTX 5090	8,192	270.3	434.2	1118.4	55.77	206.19	157	1.72
Qwen3.5-2B-IQ4_XS	IQ4_XS	Apple M4 Pro	8,192	270.3	551.0	1139.8	111.81	240.63	29	9.29
Qwen3.5-4B-IQ4_XS	IQ4_XS	NVIDIA GeForce RTX 4060 Ti	130,064	270.1	588.4	1959.4	114.04	226.52	55	4.94
Qwen3.5-0.8B-BF16	BF16	NVIDIA GB10	2,048	269.9	3229.2	3895.1	14.17	17.40	—	—
Qwen3.5-4B-Q3_K_S	Q3_K_S	NVIDIA GeForce RTX 4060 Ti	32,768	269.8	543.4	1774.6	113.80	189.64	56	4.84
Qwen3.5-4B-Q3_K_M	Q3_K_M	NVIDIA GeForce RTX 4060 Ti	16,384	269.1	547.6	1427.7	113.54	190.35	56	4.84
Qwen3.5-2B-Q6_K	Q6_K	Apple M4 Pro	130,064	268.9	665.7	3230.7	112.55	209.33	30	9.11
Qwen3.5-4B-IQ4_NL	IQ4_NL	NVIDIA GeForce RTX 4060 Ti	16,384	268.9	580.4	1860.2	113.91	193.46	56	4.84
gemma-4-26B-A4B-it-UD-IQ3_XXS	IQ3_XXS	NVIDIA GB10	32,768	267.8	1263.9	5146.6	104.10	234.57	52	5.18
Qwen3.5-4B-Q4_K_S	Q4_K_S	NVIDIA GeForce RTX 4060 Ti	16,384	267.7	456.0	1214.8	114.53	218.78	56	4.81
Qwen3.5-4B-UD-Q2_K_XL	Q2_K_XL	NVIDIA GeForce RTX 4060 Ti	16,384	267.2	750.9	3159.8	115.08	233.09	53	5.05
Qwen3.5-2B-Q4_K_M	Q4_K_M	NVIDIA GB10	130,064	267.1	3498.2	4008.2	14.61	52.39	64	4.19
Qwen3.5-4B-Q4_0	Q4_0	NVIDIA GeForce RTX 4060 Ti	16,384	266.7	666.9	1625.5	114.93	223.04	57	4.70
Qwen3.5-4B-Q4_1	Q4_1	NVIDIA GeForce RTX 4060 Ti	32,768	266.4	895.4	2686.1	114.54	214.29	56	4.75
gemma-4-26B-A4B-it-UD-IQ4_NL	IQ4_NL	NVIDIA GB10	130,064	266.2	773.8	1881.9	112.24	332.47	48	5.52
Qwen3.5-4B-Q4_K_M	Q4_K_M	NVIDIA GeForce RTX 4060 Ti	16,384	265.4	535.8	1758.4	115.47	242.33	56	4.76
Qwen3.5-2B-Q4_K_S	Q4_K_S	Apple M4 Pro	16,384	264.9	557.8	2682.7	114.67	226.06	31	8.54
Qwen3.5-2B-UD-Q4_K_XL	Q4_K_XL	Apple M4 Pro	16,384	264.8	555.7	1892.7	115.50	293.48	28	9.63
Qwen3.5-2B-BF16	BF16	Apple M4 Pro	130,064	264.5	473.8	1671.6	113.87	213.15	32	8.21
Qwen3.5-2B-Q4_K_M	Q4_K_M	Apple M4 Pro	8,192	264.2	548.9	1266.9	114.10	202.39	29	9.08
Qwen3.5-2B-Q5_K_S	Q5_K_S	Apple M4 Pro	16,384	264.1	691.7	3181.6	113.74	244.53	30	8.95
Qwen3.5-2B-Q5_K_M	Q5_K_M	Apple M4 Pro	130,064	263.9	654.4	3225.2	114.42	218.01	29	9.01
Qwen3.5-2B-UD-Q5_K_XL	Q5_K_XL	Apple M4 Pro	8,192	263.8	500.8	1111.5	114.89	212.72	30	8.94
gemma-4-26B-A4B-it-MXFP4_MOE		NVIDIA GB10	130,064	263.1	1038.5	3831.8	109.29	263.56	41	6.45
Qwen3.5-4B-UD-Q4_K_XL	Q4_K_XL	NVIDIA GeForce RTX 4060 Ti	32,768	262.0	623.0	1570.3	116.97	224.13	53	4.95
gemma-4-26B-A4B-it-UD-IQ4_XS	IQ4_XS	NVIDIA GB10	65,536	261.3	607.1	1392.0	114.86	344.41	45	5.81
Qwen3.5-2B-Q5_K_S	Q5_K_S	NVIDIA GB10	8,192	260.2	120.2	238.9	14.90	17.83	61	4.27
Qwen3.5-4B-UD-Q5_K_XL	Q5_K_XL	NVIDIA GeForce RTX 4060 Ti	32,768	257.6	635.5	1664.6	118.39	249.83	56	4.62
Qwen3.5-4B-UD-Q3_K_XL	Q3_K_XL	NVIDIA GeForce RTX 4060 Ti	130,064	257.3	475.7	1402.1	118.90	238.28	55	4.70
Qwen3.5-4B-UD-IQ3_XXS	IQ3_XXS	NVIDIA GeForce RTX 4060 Ti	130,064	257.3	694.6	1842.0	119.41	254.34	56	4.60
Qwen3.5-2B-UD-Q4_K_XL	Q4_K_XL	NVIDIA GB10	2,048	257.2	3503.9	4228.5	15.03	19.10	62	4.17
gemma-4-E4B-it-UD-IQ2_M	IQ2_M	Apple M4 Pro	16,384	257.1	1025.7	2801.7	111.74	293.50	30	8.49
Qwen3.5-2B-UD-Q5_K_XL	Q5_K_XL	NVIDIA GB10	130,064	257.0	146.8	279.7	15.04	18.01	59	4.37
Qwen3.5-4B-Q5_K_S	Q5_K_S	NVIDIA GeForce RTX 4060 Ti	32,768	256.6	539.3	1663.4	119.52	224.13	56	4.56
Qwen3.5-4B-Q6_K	Q6_K	NVIDIA GeForce RTX 4060 Ti	16,384	256.4	638.9	2286.8	119.07	227.66	56	4.59
Qwen3.5-4B-Q5_K_M	Q5_K_M	NVIDIA GeForce RTX 4060 Ti	16,384	256.1	1030.4	3185.9	118.56	201.66	58	4.39
Qwen3.5-9B-UD-IQ3_XXS	IQ3_XXS	NVIDIA GeForce RTX 5060 Ti	8,192	253.9	399.6	833.3	57.23	124.74	124	2.05
Qwen3.5-2B-Q5_K_M	Q5_K_M	NVIDIA GB10	130,064	251.2	3716.2	4188.4	15.35	54.27	63	4.02
Qwen3.5-9B-UD-IQ2_XXS	IQ2_XXS	NVIDIA GeForce RTX 4060 Ti	16,384	248.3	696.3	2251.4	124.04	222.86	58	4.25
Qwen3.5-4B-Q8_0	Q8_0	NVIDIA GeForce RTX 4060 Ti	16,384	247.7	657.8	1904.9	122.40	197.11	56	4.44
Qwen3.5-9B-UD-Q6_K_XL	Q6_K_XL	NVIDIA GeForce RTX 5060 Ti	16,384	244.8	628.4	1410.3	124.22	308.62	119	2.07
Qwen3.5-2B-Q6_K	Q6_K	NVIDIA GB10	130,064	244.4	120.2	248.6	15.66	18.76	56	4.37
Qwen3.5-27B-Q5_K_M	Q5_K_M	NVIDIA GeForce RTX 5090	8,192	243.6	1292.2	2906.3	58.49	278.24	313	0.78
Qwen3.5-27B-Q6_K	Q6_K	NVIDIA GeForce RTX 5090	8,192	242.8	945.4	1931.6	60.25	294.06	312	0.78
Qwen3.5-9B-Q3_K_M	Q3_K_M	NVIDIA GeForce RTX 4060 Ti	32,768	238.5	512.6	1296.5	129.35	280.36	60	3.95
Qwen3.5-4B-UD-Q6_K_XL	Q6_K_XL	NVIDIA GeForce RTX 4060 Ti	16,384	238.4	666.2	1608.6	128.03	264.23	57	4.21
Qwen3.5-9B-UD-IQ2_M	IQ2_M	NVIDIA GeForce RTX 4060 Ti	32,768	238.4	828.3	3132.9	128.59	258.15	61	3.93
Qwen3.6-35B-A3B-UD-IQ1_M	IQ1_M	NVIDIA GeForce RTX 5060 Ti	16,384	237.9	744.1	2151.2	126.94	307.52	110	2.16
Qwen3.5-9B-Q4_0	Q4_0	NVIDIA GeForce RTX 4060 Ti	130,064	236.7	459.5	1404.2	131.59	238.66	60	3.96
Qwen3.5-9B-IQ4_XS	IQ4_XS	NVIDIA GeForce RTX 4060 Ti	32,768	236.5	494.2	1281.0	130.12	270.36	60	3.95
Qwen3.5-9B-Q3_K_S	Q3_K_S	NVIDIA GeForce RTX 4060 Ti	32,768	236.2	510.3	1294.6	130.35	274.24	60	3.91
Qwen3.5-9B-Q4_K_S	Q4_K_S	NVIDIA GeForce RTX 4060 Ti	16,384	235.9	490.6	1273.3	130.20	271.94	60	3.91
Qwen3.5-9B-UD-Q2_K_XL	Q2_K_XL	NVIDIA GeForce RTX 4060 Ti	32,768	235.1	836.1	2497.8	130.47	285.04	61	3.83
Qwen3.5-9B-IQ4_NL	IQ4_NL	NVIDIA GeForce RTX 4060 Ti	16,384	234.7	472.0	1263.0	131.08	259.81	59	3.95
gemma-4-E4B-it-Q4_1	Q4_1	Apple M4 Pro	16,384	233.3	689.0	1898.5	121.80	336.82	33	7.07
Qwen3.5-9B-Q4_K_M	Q4_K_M	NVIDIA GeForce RTX 4060 Ti	130,064	232.8	490.2	1443.4	132.72	271.54	59	3.95
Qwen3.5-9B-UD-Q4_K_XL	Q4_K_XL	NVIDIA GeForce RTX 4060 Ti	16,384	232.2	502.2	1280.6	133.16	261.76	60	3.90
gemma-4-E4B-it-Q4_K_M	Q4_K_M	Apple M4 Pro	130,064	232.2	738.8	1901.0	122.50	331.87	32	7.35
Qwen3.5-9B-Q4_1	Q4_1	NVIDIA GeForce RTX 4060 Ti	32,768	232.1	512.1	1299.6	132.36	277.63	59	3.95
Qwen3.5-27B.Q4_K_M	Q4_K_M	NVIDIA GeForce RTX 5090	16,384	231.7	2417.4	6723.5	128.32	161.20	424	0.55
Qwen3.5-35B-A3B-UD-Q6_K_XL	Q6_K_XL	NVIDIA GeForce RTX 5090	8,192	230.3	485.9	1667.8	66.01	229.16	241	0.95
Qwen3.5-2B-Q8_0	Q8_0	NVIDIA GB10	2,048	229.6	132.4	282.7	16.77	19.74	50	4.63
Qwen3.5-27B.Q2_K	Q2_K	NVIDIA GeForce RTX 5090	8,192	227.9	1304.8	3617.7	64.56	101.70	441	0.52
gemma-4-E4B-it-IQ4_NL	IQ4_NL	Apple M4 Pro	32,768	227.0	672.6	1717.0	124.53	309.26	32	7.16
Qwen3.5-27B.Q4_K_S	Q4_K_S	NVIDIA GeForce RTX 5090	16,384	226.9	2083.3	5505.4	131.17	270.21	424	0.53
Qwen3.5-9B-UD-IQ3_XXS	IQ3_XXS	NVIDIA GeForce RTX 4060 Ti	16,384	226.3	510.0	1276.1	136.63	249.74	62	3.67
Qwen3.5-9B-UD-Q3_K_XL	Q3_K_XL	NVIDIA GeForce RTX 4060 Ti	32,768	225.8	594.7	1884.3	135.93	256.25	62	3.62
Qwen3.5-4B-UD-Q8_K_XL	Q8_K_XL	NVIDIA GeForce RTX 4060 Ti	130,064	224.8	786.7	2338.9	134.76	265.09	54	4.17
Qwen3.5-27B.Q3_K_S	Q3_K_S	NVIDIA GeForce RTX 5090	16,384	224.8	2622.3	7334.5	132.38	136.72	434	0.52
Qwen3.5-9B-UD-Q5_K_XL	Q5_K_XL	NVIDIA GeForce RTX 4060 Ti	32,768	224.2	598.4	1606.7	138.01	271.03	61	3.67
gemma-4-E4B-it-Q8_0	Q8_0	Apple M4 Pro	16,384	223.6	749.7	1812.3	123.73	358.49	34	6.58
Qwen3.5-2B-UD-Q6_K_XL	Q6_K_XL	NVIDIA GB10	8,192	223.3	4184.6	4734.8	17.39	58.86	54	4.13
gemma-4-E4B-it-Q4_K_S	Q4_K_S	Apple M4 Pro	32,768	223.2	715.8	1865.6	127.48	321.90	32	7.06
gemma-4-E4B-it-UD-IQ3_XXS	IQ3_XXS	Apple M4 Pro	130,064	221.4	892.3	2450.8	128.67	653.84	31	7.14
Qwen3.5-27B.Q3_K_M	Q3_K_M	NVIDIA GeForce RTX 5090	8,192	220.5	1419.4	3565.0	67.11	72.12	454	0.49
Qwen3.5-9B-Q5_K_S	Q5_K_S	NVIDIA GeForce RTX 4060 Ti	32,768	220.0	556.3	1406.7	139.42	265.14	62	3.54
gemma-4-E4B-it-UD-Q2_K_XL	Q2_K_XL	Apple M4 Pro	16,384	219.9	842.4	2062.2	126.81	345.92	31	7.19
Qwen3.6-27B-Q6_K	Q6_K	NVIDIA GeForce RTX 5090	8,192	219.6	774.9	1800.1	66.63	293.95	364	0.60
Qwen3.5-27B-UD-Q4_K_XL	Q4_K_XL	NVIDIA GeForce RTX 5090	32,768	217.2	1377.1	3273.2	136.91	383.77	393	0.55
Qwen3.5-9B-Q5_K_M	Q5_K_M	NVIDIA GeForce RTX 4060 Ti	16,384	216.9	483.3	1312.7	142.26	265.73	62	3.52
Qwen3.5-27B-UD-IQ3_XXS	IQ3_XXS	NVIDIA GeForce RTX 5090	32,768	216.6	937.6	2619.0	139.10	382.94	376	0.58
Qwen3.5-9B-Q6_K	Q6_K	NVIDIA GeForce RTX 4060 Ti	32,768	216.3	629.6	1582.9	142.40	285.49	63	3.46
gemma-4-E4B-it-IQ4_XS	IQ4_XS	Apple M4 Pro	130,064	215.1	708.1	1776.1	129.54	338.06	32	6.66
Qwen3.6-35B-A3B-UD-IQ2_XXS	IQ2_XXS	NVIDIA GeForce RTX 5060 Ti	16,384	214.9	1509.7	5806.7	137.40	216.85	109	1.97
gemma-4-E4B-it-UD-Q8_K_XL	Q8_K_XL	Apple M4 Pro	16,384	213.4	738.0	1869.1	128.57	364.94	34	6.29
Qwen3.5-27B-UD-Q5_K_XL	Q5_K_XL	NVIDIA GeForce RTX 5090	16,384	212.7	1096.8	3034.3	139.95	353.93	381	0.56
Qwen3.5-27B-UD-Q3_K_XL	Q3_K_XL	NVIDIA GeForce RTX 5090	32,768	212.4	923.0	2632.6	140.61	423.04	381	0.56
Qwen3.5-27B-UD-Q2_K_XL	Q2_K_XL	NVIDIA GeForce RTX 5090	130,064	211.4	1087.5	3201.3	141.71	381.30	370	0.57
gemma-4-E4B-it-UD-Q4_K_XL	Q4_K_XL	Apple M4 Pro	130,064	209.3	769.0	1944.1	134.90	339.98	32	6.60
gemma-4-E4B-it-Q6_K	Q6_K	Apple M4 Pro	32,768	206.4	769.5	1995.5	132.24	398.25	33	6.27
Qwen3.5-4B-BF16	BF16	NVIDIA GeForce RTX 4060 Ti	16,384	204.8	925.2	2692.3	146.69	281.40	57	3.58
gemma-4-26B-A4B-it-Q8_0	Q8_0	NVIDIA GB10	65,536	204.5	908.4	3043.6	145.96	481.94	40	5.18
gemma-4-E4B-it-Q5_K_S	Q5_K_S	Apple M4 Pro	130,064	204.3	801.9	1984.1	135.49	413.95	33	6.27
gemma-4-E4B-it-UD-Q3_K_XL	Q3_K_XL	Apple M4 Pro	32,768	203.4	874.6	2225.1	134.23	374.71	32	6.38
Qwen3.5-9B-Q8_0	Q8_0	NVIDIA GeForce RTX 4060 Ti	16,384	203.2	684.5	1609.3	149.41	298.37	59	3.43
gemma-4-E4B-it-UD-Q6_K_XL	Q6_K_XL	Apple M4 Pro	130,064	201.9	801.6	1940.5	137.42	398.91	33	6.19
gemma-4-E4B-it-Q3_K_M	Q3_K_M	Apple M4 Pro	32,768	200.0	833.2	1942.2	136.27	382.00	32	6.23
gemma-4-E4B-it-BF16	BF16	Apple M4 Pro	32,768	199.0	814.5	1941.9	135.18	390.12	36	5.47
Qwen3.5-9B-UD-Q6_K_XL	Q6_K_XL	NVIDIA GeForce RTX 4060 Ti	16,384	196.8	682.8	1658.9	157.66	314.24	64	3.06
gemma-4-E4B-it-Q5_K_M	Q5_K_M	Apple M4 Pro	16,384	196.0	773.5	2053.9	141.14	421.28	33	5.97
gemma-4-E4B-it-UD-Q5_K_XL	Q5_K_XL	Apple M4 Pro	32,768	189.0	868.2	2202.7	145.51	427.08	32	5.89
Qwen3.6-35B-A3B-UD-IQ1_M	IQ1_M	NVIDIA GeForce RTX 4060 Ti	130,064	188.1	637.3	1480.7	163.67	338.87	61	3.07
Qwen3.6-35B-A3B-UD-IQ1_M	IQ1_M	NVIDIA GB10	32,768	186.9	665.8	1807.7	163.55	373.03	50	3.78
gemma-4-E4B-it-Q3_K_S	Q3_K_S	Apple M4 Pro	130,064	186.7	921.8	2299.5	142.25	382.12	32	5.85
Qwen3.5-2B-UD-Q8_K_XL	Q8_K_XL	NVIDIA GB10	32,768	186.4	132.3	249.8	20.44	23.77	44	4.23
Qwen3.6-35B-A3B-UD-IQ2_XXS	IQ2_XXS	NVIDIA GeForce RTX 4060 Ti	32,768	186.1	628.2	1397.1	165.86	341.65	63	2.98
Qwen3.6-35B-A3B-UD-IQ2_M	IQ2_M	NVIDIA GeForce RTX 4060 Ti	16,384	185.9	657.6	1452.5	164.37	365.50	62	2.99
Qwen3.6-35B-A3B-UD-Q4_K_S	Q4_K_S	NVIDIA GeForce RTX 4060 Ti	16,384	185.0	806.0	1913.3	163.88	330.62	60	3.07
Qwen3.6-35B-A3B-UD-IQ2_M	IQ2_M	NVIDIA GB10	65,536	184.2	811.1	1686.3	163.98	378.97	53	3.46
Qwen3.6-35B-A3B-UD-Q4_K_M	Q4_K_M	NVIDIA GeForce RTX 4060 Ti	16,384	183.2	911.9	2600.1	166.34	369.06	61	3.00
Qwen3.6-35B-A3B-UD-Q4_K_XL	Q4_K_XL	NVIDIA GeForce RTX 4060 Ti	32,768	182.7	685.2	1927.1	168.90	379.66	60	3.05
Qwen3.6-35B-A3B-UD-IQ2_XXS	IQ2_XXS	NVIDIA GB10	65,536	182.7	770.1	1543.3	166.13	416.75	54	3.40
Qwen3.6-35B-A3B-UD-Q3_K_M	Q3_K_M	NVIDIA GeForce RTX 4060 Ti	32,768	182.1	746.4	1929.5	168.48	343.16	61	3.00
Qwen3.6-35B-A3B-UD-Q2_K_XL	Q2_K_XL	NVIDIA GeForce RTX 4060 Ti	32,768	181.5	670.6	1446.2	168.81	355.93	63	2.89
Qwen3.6-35B-A3B-UD-IQ3_XXS	IQ3_XXS	NVIDIA GeForce RTX 4060 Ti	130,064	181.4	773.5	2103.7	170.05	356.00	63	2.87
Qwen3.6-35B-A3B-UD-Q3_K_XL	Q3_K_XL	NVIDIA GeForce RTX 4060 Ti	130,064	180.5	722.2	1653.2	168.59	378.16	61	2.97
Qwen3.6-35B-A3B-Q8_0	Q8_0	NVIDIA GeForce RTX 5090	8,192	180.5	672.1	1941.0	82.78	340.00	145	1.25
Qwen3.6-35B-A3B-UD-IQ4_NL_XL	IQ4_NL_XL	NVIDIA GeForce RTX 4060 Ti	32,768	180.3	766.9	2134.3	170.34	345.29	61	2.97
Qwen3.6-35B-A3B-MXFP4_MOE		NVIDIA GeForce RTX 4060 Ti	130,064	179.7	1048.6	3215.1	168.61	379.27	61	2.96
Qwen3.5-9B-UD-Q8_K_XL	Q8_K_XL	NVIDIA GeForce RTX 4060 Ti	16,384	179.7	612.0	1274.9	172.03	337.52	60	3.01
Qwen3.6-35B-A3B-UD-IQ2_M	IQ2_M	NVIDIA GeForce RTX 5060 Ti	8,192	179.5	746.4	14514.2	82.66	195.51	95	1.89
Qwen3.6-35B-A3B-UD-Q3_K_S	Q3_K_S	NVIDIA GeForce RTX 4060 Ti	16,384	179.3	667.9	1519.2	170.77	344.76	62	2.91
Qwen3.6-35B-A3B-UD-IQ3_S	IQ3_S	NVIDIA GeForce RTX 4060 Ti	130,064	178.5	650.9	1628.9	171.01	385.86	65	2.76
Qwen3.6-35B-A3B-UD-Q5_K_S	Q5_K_S	NVIDIA GeForce RTX 4060 Ti	32,768	178.2	770.1	2244.6	171.65	382.30	61	2.95
Qwen3.6-35B-A3B-UD-IQ4_NL	IQ4_NL	NVIDIA GeForce RTX 4060 Ti	32,768	178.0	858.3	2623.8	171.82	374.43	61	2.91
Qwen3.6-35B-A3B-UD-Q2_K_XL	Q2_K_XL	NVIDIA GB10	32,768	177.7	1221.7	4127.7	168.56	325.02	54	3.30
Qwen3.6-35B-A3B-UD-IQ4_XS	IQ4_XS	NVIDIA GeForce RTX 4060 Ti	16,384	177.5	776.6	2095.9	171.62	334.71	61	2.89
Qwen3.6-35B-A3B-UD-IQ3_XXS	IQ3_XXS	NVIDIA GB10	130,064	174.6	1002.3	3597.4	173.23	406.04	52	3.38
Qwen3.5-9B-UD-Q8_K_XL	Q8_K_XL	NVIDIA GeForce RTX 5060 Ti	8,192	174.5	416.4	763.3	87.56	218.65	112	1.55
Qwen3.6-35B-A3B-UD-Q3_K_S	Q3_K_S	NVIDIA GB10	32,768	174.0	988.6	3132.7	173.78	393.40	49	3.52
Qwen3.6-35B-A3B-APEX-Compact		NVIDIA GB10	130,064	173.6	1167.9	3658.8	173.68	410.94	52	3.31
Qwen3.6-35B-A3B-APEX-I-Mini		NVIDIA GB10	65,536	172.8	883.4	2585.8	174.22	403.83	51	3.38
Qwen3.6-35B-A3B-MXFP4_MOE		NVIDIA GB10	32,768	172.7	1148.3	3648.6	173.74	389.71	46	3.79
Qwen3.6-35B-A3B-APEX-I-Compact		NVIDIA GB10	32,768	172.3	1025.8	3289.6	174.29	407.67	51	3.40
Qwen3.6-35B-A3B-UD-Q3_K_M	Q3_K_M	NVIDIA GB10	32,768	172.1	827.7	2071.1	176.50	402.27	51	3.39
Qwen3.6-35B-A3B-UD-IQ3_S	IQ3_S	NVIDIA GB10	32,768	170.8	811.6	1697.8	177.53	411.92	55	3.13
Qwen3.6-35B-A3B-UD-Q4_K_S	Q4_K_S	NVIDIA GB10	32,768	169.9	833.3	1984.4	178.12	405.99	47	3.64
Qwen3.6-35B-A3B-UD-Q3_K_XL	Q3_K_XL	NVIDIA GB10	32,768	169.5	989.2	3113.1	177.47	401.13	49	3.46
Qwen3.6-35B-A3B-UD-IQ4_NL_XL	IQ4_NL_XL	NVIDIA GB10	65,536	168.7	896.5	1880.5	177.84	405.94	49	3.42
Qwen3.6-35B-A3B-UD-IQ4_XS	IQ4_XS	NVIDIA GB10	32,768	168.5	1085.0	3928.4	178.01	393.31	49	3.44
Qwen3.6-35B-A3B-UD-Q4_K_XL	Q4_K_XL	NVIDIA GB10	65,536	168.2	883.5	1784.4	180.94	433.75	48	3.52
Qwen3.6-35B-A3B-UD-Q4_K_M	Q4_K_M	NVIDIA GB10	32,768	166.7	860.1	1794.3	182.26	415.16	49	3.44
Qwen3.5-9B-UD-IQ2_M	IQ2_M	NVIDIA GeForce RTX 5060 Ti	130,064	165.9	459.7	1356.9	138.09	297.09	148	1.12
Qwen3.6-35B-A3B-APEX-I-Quality	Quality	NVIDIA GB10	130,064	165.6	1066.9	3407.9	183.30	436.74	51	3.23
Qwen3.6-35B-A3B-UD-IQ4_NL	IQ4_NL	NVIDIA GB10	65,536	164.2	842.7	2067.8	182.37	419.76	50	3.32
Qwen3.6-35B-A3B-APEX-Quality	Quality	NVIDIA GB10	65,536	164.0	909.7	2343.2	184.77	430.85	51	3.20
Qwen3.5-35B-A3B-Q8_0	Q8_0	NVIDIA GeForce RTX 5090	32,768	162.5	969.5	3696.2	93.20	183.07	177	0.92
Qwen3.6-35B-A3B-APEX-Balanced	Balanced	NVIDIA GB10	65,536	160.8	1045.7	2827.2	187.71	437.93	51	3.13
Qwen3.5-2B-BF16	BF16	NVIDIA GB10	65,536	160.3	153.7	287.1	24.18	27.15	41	3.94
Qwen3.6-35B-A3B-UD-Q5_K_S	Q5_K_S	NVIDIA GB10	130,064	158.4	942.9	2562.1	190.03	433.30	46	3.44
Qwen3.6-35B-A3B-APEX-I-Balanced	Balanced	NVIDIA GB10	65,536	157.3	1179.8	4036.4	189.87	437.54	51	3.10
Qwen3.6-35B-A3B-UD-Q5_K_XL	Q5_K_XL	NVIDIA GB10	65,536	156.3	884.1	2286.9	193.84	442.09	47	3.32
Qwen3.6-35B-A3B-UD-Q5_K_M	Q5_K_M	NVIDIA GB10	65,536	155.0	1109.2	3952.2	192.39	444.24	47	3.32
Qwen3.5-9B-BF16	BF16	NVIDIA GeForce RTX 4060 Ti	16,384	155.0	709.9	1322.1	197.24	399.90	60	2.58
Qwen3.6-35B-A3B-UD-Q5_K_XL	Q5_K_XL	NVIDIA GeForce RTX 4060 Ti	8,192	152.4	507.0	1075.2	100.09	183.41	60	2.53
Qwen3.6-35B-A3B-UD-Q5_K_M	Q5_K_M	NVIDIA GeForce RTX 4060 Ti	8,192	151.8	526.0	1185.0	100.02	209.87	61	2.51
Qwen3.5-0.8B-Q8_0	Q8_0	Apple M4 Pro	8,192	151.5	42495.8	49174.5	26.20	47.55	—	—
Qwen3.5-4B-UD-IQ2_XXS	IQ2_XXS	NVIDIA GB10	32,768	150.1	266.0	568.0	25.71	30.09	69	2.18
Qwen3.5-0.8B-UD-Q8_K_XL	Q8_K_XL	Apple M4 Pro	8,192	149.8	42573.8	49894.0	26.47	48.96	—	—
Qwen3.5-27B-UD-IQ2_XXS	IQ2_XXS	NVIDIA GeForce RTX 5090	32,768	149.0	529.0	853.4	24.88	30.57	370	0.40
Qwen3.6-35B-A3B-UD-Q6_K	Q6_K	NVIDIA GB10	65,536	148.5	1188.0	3671.6	201.09	469.03	48	3.07
gemma-4-26B-A4B-it-UD-IQ2_XXS	IQ2_XXS	Apple M4 Pro	32,768	148.2	2422.7	15602.3	203.08	422.19	31	4.86
Qwen3.5-0.8B-UD-IQ2_XXS	IQ2_XXS	Apple M4 Pro	32,768	146.9	118.2	411.2	26.86	29.21	—	—
Qwen3.5-0.8B-UD-IQ3_XXS	IQ3_XXS	Apple M4 Pro	32,768	146.8	118.0	405.3	26.88	28.99	—	—
Qwen3.5-0.8B-IQ4_NL	IQ4_NL	Apple M4 Pro	16,384	146.0	119.6	409.1	26.88	29.07	—	—
Qwen3.6-35B-A3B-UD-Q6_K_XL	Q6_K_XL	NVIDIA GB10	65,536	146.0	1184.8	3345.7	204.44	473.44	47	3.12
gemma-4-26B-A4B-it-UD-IQ2_M	IQ2_M	Apple M4 Pro	130,064	146.0	2321.6	13433.6	203.12	386.16	32	4.62
Qwen3.5-0.8B-Q4_1	Q4_1	Apple M4 Pro	8,192	144.8	117.2	407.1	27.26	29.32	—	—
Qwen3.5-0.8B-UD-IQ2_M	IQ2_M	Apple M4 Pro	8,192	144.0	6872.1	7849.1	27.42	29.74	26	5.52
Qwen3.5-0.8B-Q4_0	Q4_0	Apple M4 Pro	32,768	144.0	19735.0	22433.1	27.50	47.19	—	—
Qwen3.5-9B-UD-IQ2_XXS	IQ2_XXS	NVIDIA GB10	130,064	143.4	607.8	1495.5	215.76	386.47	54	2.67
Qwen3.5-0.8B-UD-Q2_K_XL	Q2_K_XL	Apple M4 Pro	130,064	143.0	119.9	415.4	27.60	29.68	—	—
Qwen3.5-4B-UD-IQ2_M	IQ2_M	NVIDIA GB10	65,536	142.3	263.0	600.6	26.78	31.44	70	2.04
Qwen3.6-35B-A3B-Q8_0	Q8_0	NVIDIA GB10	65,536	141.7	1125.3	3109.1	210.87	498.27	44	3.20
Qwen3.5-0.8B-IQ4_XS	IQ4_XS	Apple M4 Pro	130,064	141.4	106.7	346.6	27.71	29.83	—	—
Qwen3.5-0.8B-UD-Q6_K_XL	Q6_K_XL	Apple M4 Pro	16,384	141.4	112.7	366.6	27.27	29.65	—	—
gemma-4-26B-A4B-it-UD-Q3_K_M	Q3_K_M	Apple M4 Pro	32,768	141.4	1747.8	9828.5	212.13	505.26	32	4.38
Qwen3.5-9B-UD-IQ2_M	IQ2_M	NVIDIA GB10	2,048	141.3	837.4	1476.6	216.46	321.79	56	2.53
Qwen3.5-0.8B-Q4_K_S	Q4_K_S	Apple M4 Pro	8,192	141.2	115.9	342.7	27.42	29.55	—	—
Qwen3.5-9B-IQ4_XS	IQ4_XS	NVIDIA GB10	2,048	140.9	625.9	1447.2	220.36	391.25	54	2.60
Qwen3.5-9B-Q4_0	Q4_0	NVIDIA GB10	8,192	140.4	645.1	1469.9	219.86	421.54	54	2.61
Qwen3.5-0.8B-UD-Q3_K_XL	Q3_K_XL	Apple M4 Pro	16,384	140.4	19989.5	22869.5	28.16	30.39	—	—
Qwen3.5-9B-Q3_K_S	Q3_K_S	NVIDIA GB10	2,048	139.8	722.4	1910.5	220.24	393.59	56	2.49
Qwen3.5-9B-Q3_K_M	Q3_K_M	NVIDIA GB10	2,048	139.8	610.6	1472.5	221.07	445.21	55	2.55
Qwen3.5-0.8B-UD-Q4_K_XL	Q4_K_XL	Apple M4 Pro	8,192	139.8	94.2	241.5	27.86	30.38	—	—
Qwen3.5-9B-Q4_1	Q4_1	NVIDIA GB10	2,048	139.4	609.5	1296.0	221.53	406.28	52	2.68
Qwen3.5-9B-IQ4_NL	IQ4_NL	NVIDIA GB10	8,192	139.3	647.6	1457.3	220.74	411.77	55	2.55
Qwen3.5-0.8B-Q3_K_M	Q3_K_M	Apple M4 Pro	8,192	139.3	6756.8	8028.9	28.10	48.73	—	—
Qwen3.5-27B-UD-IQ2_M	IQ2_M	NVIDIA GeForce RTX 5090	16,384	139.3	46973.5	52639.8	27.85	164.10	408	0.34
Qwen3.6-35B-A3B-UD-Q8_K_XL	Q8_K_XL	NVIDIA GB10	65,536	138.9	1169.8	4154.2	214.65	498.22	43	3.25
gemma-4-26B-A4B-it-UD-Q4_K_S	Q4_K_S	Apple M4 Pro	16,384	138.8	1940.5	10591.8	216.85	500.12	32	4.38
gemma-4-26B-A4B-it-UD-Q3_K_XL	Q3_K_XL	Apple M4 Pro	130,064	138.7	1755.9	9140.4	216.48	637.48	33	4.27
Qwen3.5-0.8B-Q4_K_M	Q4_K_M	Apple M4 Pro	130,064	138.5	46912.2	53959.1	28.60	47.80	—	—
Qwen3.5-9B-UD-Q2_K_XL	Q2_K_XL	NVIDIA GB10	32,768	138.3	903.3	3314.7	223.26	388.25	55	2.53
gemma-4-26B-A4B-it-UD-Q4_K_M	Q4_K_M	Apple M4 Pro	130,064	138.3	1120.9	4029.3	219.70	740.99	34	4.08
Qwen3.5-9B-Q4_K_S	Q4_K_S	NVIDIA GB10	2,048	138.2	776.9	2060.6	221.14	455.83	55	2.51
Qwen3.5-4B-UD-IQ3_XXS	IQ3_XXS	NVIDIA GB10	8,192	137.2	6850.7	7897.4	28.33	33.42	72	1.91
gemma-4-26B-A4B-it-UD-IQ3_XXS	IQ3_XXS	Apple M4 Pro	130,064	136.8	1901.3	10078.3	215.73	529.76	32	4.30
Qwen3.5-0.8B-Q5_K_S	Q5_K_S	Apple M4 Pro	16,384	136.4	47723.9	54522.5	29.11	50.06	—	—
Qwen3.5-9B-Q4_K_M	Q4_K_M	NVIDIA GB10	2,048	136.2	672.0	1488.8	225.51	430.90	54	2.53
Qwen3.5-0.8B-Q3_K_S	Q3_K_S	Apple M4 Pro	130,064	136.0	122.9	424.2	29.04	31.15	—	—
Qwen3.5-0.8B-Q5_K_M	Q5_K_M	Apple M4 Pro	8,192	135.2	98.8	245.2	28.89	31.18	—	—
gemma-4-26B-A4B-it-UD-IQ4_NL	IQ4_NL	Apple M4 Pro	130,064	135.0	1070.0	3874.7	221.06	720.57	32	4.23
gemma-4-26B-A4B-it-UD-Q4_K_XL	Q4_K_XL	Apple M4 Pro	16,384	134.9	1450.8	7524.6	223.81	675.05	31	4.31
Qwen3.5-4B-UD-Q2_K_XL	Q2_K_XL	NVIDIA GB10	130,064	134.8	7055.9	8040.8	28.69	33.93	70	1.91
Qwen3.5-9B-UD-IQ3_XXS	IQ3_XXS	NVIDIA GB10	65,536	134.5	712.6	2137.5	230.06	420.25	56	2.42
Qwen3.5-9B-UD-Q3_K_XL	Q3_K_XL	NVIDIA GB10	130,064	134.5	681.9	2237.6	228.86	423.40	55	2.47
gemma-4-26B-A4B-it-MXFP4_MOE		Apple M4 Pro	130,064	134.3	1795.7	10368.5	215.53	513.84	33	4.02
gemma-4-26B-A4B-it-UD-Q2_K_XL	Q2_K_XL	Apple M4 Pro	16,384	134.3	1235.4	4891.6	224.53	723.87	31	4.33
Qwen3.5-0.8B-Q6_K	Q6_K	Apple M4 Pro	8,192	134.2	48291.2	55414.2	29.58	49.11	—	—
Qwen3.5-0.8B-UD-Q5_K_XL	Q5_K_XL	Apple M4 Pro	130,064	134.0	7011.0	8085.4	29.32	31.40	—	—
Qwen3.5-4B-Q4_0	Q4_0	NVIDIA GB10	130,064	133.8	194.3	336.1	29.08	41.59	59	2.26
Qwen3.5-9B-UD-Q4_K_XL	Q4_K_XL	NVIDIA GB10	130,064	133.8	641.6	1766.6	229.11	429.58	53	2.53
gemma-4-26B-A4B-it-UD-IQ4_XS	IQ4_XS	Apple M4 Pro	16,384	132.9	958.9	2118.6	223.10	745.25	32	4.14
Qwen3.5-4B-IQ4_NL	IQ4_NL	NVIDIA GB10	8,192	132.9	7132.1	8047.0	29.42	95.31	64	2.08
gemma-4-26B-A4B-it-UD-IQ3_S	IQ3_S	Apple M4 Pro	130,064	132.6	1878.0	9793.5	219.79	585.76	32	4.14
Qwen3.5-4B-Q4_1	Q4_1	NVIDIA GB10	8,192	132.1	7348.0	8227.9	29.45	33.88	62	2.12
gemma-4-26B-A4B-it-Q8_0	Q8_0	Apple M4 Pro	32,768	131.9	1157.9	2880.3	230.64	782.27	34	3.88
Qwen3.5-9B-Q5_K_M	Q5_K_M	NVIDIA GB10	2,048	130.9	681.5	1554.3	233.66	431.81	54	2.41
Qwen3.5-4B-IQ4_XS	IQ4_XS	NVIDIA GB10	65,536	130.4	7197.3	8260.1	29.95	88.78	62	2.09
Qwen3.5-9B-UD-Q5_K_XL	Q5_K_XL	NVIDIA GB10	65,536	130.4	713.5	1948.8	235.10	496.52	54	2.41
Qwen3.5-9B-Q5_K_S	Q5_K_S	NVIDIA GB10	65,536	130.1	717.5	1894.9	237.05	481.74	54	2.40
Qwen3.5-4B-Q3_K_S	Q3_K_S	NVIDIA GB10	8,192	129.9	7232.2	8279.0	29.95	35.74	70	1.86
gemma-4-26B-A4B-it-UD-Q6_K_XL	Q6_K_XL	Apple M4 Pro	16,384	129.8	1216.4	4131.1	236.66	794.69	33	3.99
Qwen3.5-4B-Q3_K_M	Q3_K_M	NVIDIA GB10	65,536	129.0	219.0	504.4	29.91	34.52	67	1.92
Qwen3.5-9B-Q6_K	Q6_K	NVIDIA GB10	2,048	128.5	685.7	1326.4	237.46	482.46	54	2.40
Qwen3.5-4B-UD-Q3_K_XL	Q3_K_XL	NVIDIA GB10	2,048	127.8	258.5	584.7	30.21	34.25	68	1.87
gemma-4-26B-A4B-it-UD-Q6_K	Q6_K	Apple M4 Pro	130,064	127.7	1333.0	4945.6	239.58	807.02	32	3.98
gemma-4-26B-A4B-it-UD-Q5_K_XL	Q5_K_XL	Apple M4 Pro	32,768	126.2	1383.3	4340.6	241.24	825.33	35	3.64
Qwen3.5-4B-Q4_K_S	Q4_K_S	NVIDIA GB10	32,768	125.8	7239.3	8548.3	31.02	73.27	72	1.74
gemma-4-26B-A4B-it-UD-Q5_K_S	Q5_K_S	Apple M4 Pro	32,768	125.3	1136.6	3603.2	244.55	815.21	32	3.89
Qwen3.5-9B-UD-Q3_K_XL	Q3_K_XL	NVIDIA GeForce RTX 5060 Ti	16,384	125.2	448.8	1322.5	141.75	313.99	150	0.84
Qwen3.5-0.8B-BF16	BF16	Apple M4 Pro	8,192	124.0	52224.8	59107.6	32.02	53.91	—	—
Qwen3.5-9B-Q8_0	Q8_0	NVIDIA GB10	2,048	123.9	845.7	2047.5	244.01	482.00	52	2.36
Qwen3.5-4B-Q4_K_M	Q4_K_M	NVIDIA GB10	32,768	122.5	22965.2	25664.7	32.15	37.88	72	1.71
gemma-4-26B-A4B-it-UD-Q5_K_M	Q5_K_M	Apple M4 Pro	16,384	121.1	1487.7	6119.2	247.70	833.04	32	3.77
Qwen3.5-4B-Q5_K_S	Q5_K_S	NVIDIA GB10	2,048	120.1	7770.2	8913.3	32.62	99.59	72	1.67
Qwen3.5-4B-UD-Q4_K_XL	Q4_K_XL	NVIDIA GB10	8,192	119.9	7842.4	8875.8	32.27	100.00	69	1.74
gemma-4-26B-A4B-it-UD-Q8_K_XL	Q8_K_XL	Apple M4 Pro	32,768	119.6	1219.9	2314.6	256.23	851.36	30	4.03
Qwen3.5-4B-Q5_K_M	Q5_K_M	NVIDIA GB10	2,048	118.7	8175.7	9068.8	32.47	36.62	71	1.66
Qwen3.5-4B-UD-Q5_K_XL	Q5_K_XL	NVIDIA GB10	130,064	116.0	8251.0	9253.3	33.60	104.30	68	1.70
Qwen3.6-35B-A3B-UD-Q6_K	Q6_K	NVIDIA GeForce RTX 4060 Ti	8,192	113.7	768.9	2171.1	132.59	289.80	54	2.11
Qwen3.5-4B-Q6_K	Q6_K	NVIDIA GB10	8,192	111.9	8514.4	9515.3	34.79	39.73	66	1.70
Qwen3.5-27B-Q8_0	Q8_0	NVIDIA GeForce RTX 5090	8,192	106.1	1066.0	2154.8	142.24	462.01	168	0.63
Qwen3.5-4B-Q8_0	Q8_0	NVIDIA GB10	2,048	105.0	261.9	631.8	36.98	41.31	55	1.90
Qwen3.5-4B-UD-Q6_K_XL	Q6_K_XL	NVIDIA GB10	32,768	103.3	228.6	385.8	37.69	57.61	60	1.72
Qwen3.5-9B-UD-IQ2_XXS	IQ2_XXS	Apple M4 Pro	8,192	102.7	2276.4	6056.8	290.91	592.74	33	3.09
Qwen3.5-9B-Q8_0	Q8_0	Apple M4 Pro	8,192	102.0	1488.6	3216.9	295.60	785.92	36	2.86
Qwen3.5-9B-UD-Q2_K_XL	Q2_K_XL	NVIDIA GeForce RTX 5060 Ti	130,064	102.0	499.5	1391.4	166.73	406.59	151	0.68
Qwen3.5-9B-Q3_K_M	Q3_K_M	Apple M4 Pro	8,192	100.0	1768.0	3606.8	303.67	959.69	33	2.99
Qwen3.5-35B-A3B-UD-Q8_K_XL	Q8_K_XL	NVIDIA GeForce RTX 5090	8,192	99.3	1493.8	7358.5	152.26	266.40	140	0.71
Qwen3.6-35B-A3B-APEX-I-Compact		NVIDIA GeForce RTX 5090	32,768	98.7	200.1	467.7	9.35	12.71	159	0.62
Qwen3.6-35B-A3B-APEX-Compact		NVIDIA GeForce RTX 5090	32,768	98.6	200.6	467.6	9.39	12.56	159	0.62
Qwen3.5-9B-IQ4_NL	IQ4_NL	Apple M4 Pro	8,192	98.6	1501.2	3516.8	311.15	727.10	34	2.87
Qwen3.5-9B-UD-IQ3_XXS	IQ3_XXS	Apple M4 Pro	32,768	98.3	1915.6	6182.4	310.77	1069.33	33	3.00
Qwen3.5-9B-Q4_1	Q4_1	Apple M4 Pro	16,384	97.9	1533.3	4295.7	315.69	921.00	34	2.85
Qwen3.5-9B-Q4_0	Q4_0	Apple M4 Pro	8,192	97.0	1444.8	3339.3	316.65	723.35	35	2.80
Qwen3.5-9B-Q4_K_S	Q4_K_S	Apple M4 Pro	8,192	97.0	1748.3	3977.9	313.36	776.21	35	2.80
Qwen3.5-9B-UD-Q4_K_XL	Q4_K_XL	Apple M4 Pro	8,192	96.6	1868.7	4002.7	312.44	907.41	34	2.88
Qwen3.6-35B-A3B-APEX-I-Mini		NVIDIA GeForce RTX 5090	32,768	96.5	197.7	406.3	9.63	12.45	154	0.63
Qwen3.5-9B-UD-Q8_K_XL	Q8_K_XL	Apple M4 Pro	32,768	96.3	1507.0	3529.3	314.65	994.82	36	2.66
Qwen3.5-9B-Q6_K	Q6_K	Apple M4 Pro	8,192	96.2	1657.1	3401.7	317.58	853.15	35	2.72
Qwen3.5-9B-UD-Q6_K_XL	Q6_K_XL	Apple M4 Pro	8,192	95.9	1533.0	3547.3	316.34	811.18	35	2.71
Qwen3.5-9B-IQ4_XS	IQ4_XS	Apple M4 Pro	8,192	95.8	1575.1	3298.7	319.50	860.72	34	2.81
Qwen3.5-9B-UD-IQ2_M	IQ2_M	Apple M4 Pro	130,064	95.6	1980.3	8685.5	317.91	1027.83	33	2.86
Qwen3.5-9B-UD-Q3_K_XL	Q3_K_XL	Apple M4 Pro	8,192	95.4	1729.6	3604.4	316.21	838.64	34	2.84
Qwen3.5-9B-BF16	BF16	Apple M4 Pro	130,064	95.2	1757.2	6149.1	317.06	993.41	37	2.55
Qwen3.5-9B-Q3_K_S	Q3_K_S	Apple M4 Pro	8,192	95.0	1827.9	4503.2	316.06	938.13	33	2.85
Qwen3.5-9B-UD-Q6_K_XL	Q6_K_XL	NVIDIA GB10	32,768	94.8	905.0	2354.9	326.09	569.35	47	2.00
Qwen3.5-9B-UD-Q5_K_XL	Q5_K_XL	Apple M4 Pro	16,384	94.8	1792.9	5849.1	321.74	1133.23	35	2.71
Qwen3.5-9B-UD-Q2_K_XL	Q2_K_XL	Apple M4 Pro	8,192	94.5	1986.3	5022.6	319.19	784.08	33	2.83
Qwen3.5-9B-Q5_K_M	Q5_K_M	Apple M4 Pro	8,192	93.3	1767.4	3987.6	327.35	850.66	34	2.75
Qwen3.5-9B-Q5_K_S	Q5_K_S	Apple M4 Pro	8,192	93.2	1803.4	3868.8	328.35	991.41	33	2.82
Qwen3.5-9B-Q4_K_M	Q4_K_M	Apple M4 Pro	32,768	92.8	1718.6	6038.4	331.24	1127.39	34	2.77
Qwen3.6-35B-A3B-UD-Q6_K_XL	Q6_K_XL	NVIDIA GeForce RTX 4060 Ti	8,192	92.5	810.9	2237.2	163.72	528.50	48	1.95
Qwen3.5-9B-UD-Q8_K_XL	Q8_K_XL	NVIDIA GB10	32,768	90.0	986.3	2314.6	344.00	719.49	47	1.90
Qwen3.5-4B-UD-Q8_K_XL	Q8_K_XL	NVIDIA GB10	130,064	88.0	10487.0	12018.8	44.39	121.85	51	1.73
Qwen3.6-35B-A3B-UD-IQ4_NL	IQ4_NL	Apple M4 Pro	8,192	86.6	895.9	3428.6	175.22	278.37	36	2.43
Qwen3.6-35B-A3B-UD-Q3_K_M	Q3_K_M	Apple M4 Pro	8,192	86.3	776.7	1931.7	177.34	361.02	35	2.48
Qwen3.6-35B-A3B-UD-IQ4_NL_XL	IQ4_NL_XL	Apple M4 Pro	8,192	85.5	725.0	1753.7	179.65	385.76	36	2.41
Qwen3.6-35B-A3B-UD-Q3_K_XL	Q3_K_XL	Apple M4 Pro	8,192	85.4	978.4	3153.6	177.65	385.27	35	2.46
Qwen3.6-35B-A3B-UD-IQ4_XS	IQ4_XS	Apple M4 Pro	8,192	85.3	782.8	2088.6	177.23	300.55	36	2.40
Qwen3.6-35B-A3B-UD-Q4_K_S	Q4_K_S	Apple M4 Pro	8,192	84.9	830.5	2156.0	177.51	292.29	35	2.43
Qwen3.6-27B-IQ4_XS	IQ4_XS	NVIDIA GB10	32,768	84.6	2182.6	5730.7	351.24	871.88	56	1.51
Qwen3.6-35B-A3B-UD-Q3_K_S	Q3_K_S	Apple M4 Pro	8,192	84.3	702.3	2068.2	183.34	416.05	35	2.39
Qwen3.6-35B-A3B-APEX-Quality	Quality	NVIDIA GeForce RTX 5090	32,768	84.2	232.1	581.0	11.01	14.58	152	0.55
Qwen3.6-35B-A3B-APEX-I-Quality	Quality	NVIDIA GeForce RTX 5090	8,192	84.2	231.5	575.4	11.02	14.67	152	0.56
Qwen3.6-35B-A3B-UD-IQ2_M	IQ2_M	Apple M4 Pro	8,192	84.0	1008.4	4504.0	182.36	363.62	34	2.50
Qwen3.6-35B-A3B-UD-IQ2_XXS	IQ2_XXS	Apple M4 Pro	8,192	83.2	770.5	2187.0	185.21	473.13	33	2.51
Qwen3.6-35B-A3B-UD-IQ3_XXS	IQ3_XXS	Apple M4 Pro	8,192	83.0	1079.7	3460.1	182.93	416.71	34	2.46
Qwen3.5-9B-BF16	BF16	NVIDIA GB10	65,536	82.7	959.6	1997.9	369.07	760.01	46	1.79
Qwen3.6-35B-A3B-UD-IQ1_M	IQ1_M	Apple M4 Pro	8,192	82.3	890.1	3093.4	186.29	455.39	34	2.44
Qwen3.6-35B-A3B-UD-IQ3_S	IQ3_S	Apple M4 Pro	8,192	82.1	853.6	2260.5	188.43	467.04	35	2.37
Qwen3.6-35B-A3B-UD-Q2_K_XL	Q2_K_XL	Apple M4 Pro	32,768	81.6	1154.8	2142.9	380.33	881.87	34	2.42
Qwen3.6-27B-Q3_K_S	Q3_K_S	NVIDIA GB10	65,536	81.4	2445.2	8520.1	362.65	888.33	61	1.33
Qwen3.6-27B-Q2_K	Q2_K	NVIDIA GB10	65,536	81.1	2006.5	4659.3	353.68	893.19	62	1.31
Qwen3.6-27B-Q3_K_L	Q3_K_L	NVIDIA GB10	65,536	80.4	2699.5	10574.5	368.90	952.70	61	1.32
Qwen3.6-27B-Q3_K_M	Q3_K_M	NVIDIA GB10	65,536	80.3	2890.2	10165.3	366.87	1092.69	61	1.32
Qwen3.6-35B-A3B-UD-Q6_K	Q6_K	Apple M4 Pro	8,192	79.9	1038.0	3067.0	190.64	383.86	36	2.21
Qwen3.6-35B-A3B-MXFP4_MOE		Apple M4 Pro	130,064	79.6	1425.8	5483.3	385.23	882.65	35	2.26
Qwen3.6-27B-Q4_K_S	Q4_K_S	NVIDIA GB10	130,064	79.6	2343.9	7674.4	363.45	872.45	54	1.46
Qwen3.6-35B-A3B-UD-Q4_K_XL	Q4_K_XL	Apple M4 Pro	130,064	79.5	1555.1	5484.8	385.70	891.42	35	2.30
Qwen3.6-35B-A3B-UD-Q4_K_M	Q4_K_M	Apple M4 Pro	8,192	79.4	1045.7	3042.3	191.23	339.58	35	2.28
Qwen3.6-35B-A3B-Q8_0	Q8_0	Apple M4 Pro	16,384	78.6	1652.2	4598.5	389.37	916.73	35	2.28
Qwen3.6-35B-A3B-UD-Q6_K_XL	Q6_K_XL	Apple M4 Pro	16,384	78.5	1624.1	4927.4	390.48	926.09	35	2.22
Qwen3.6-35B-A3B-APEX-Balanced	Balanced	NVIDIA GeForce RTX 5090	32,768	78.5	250.6	626.8	11.81	15.21	150	0.52
Qwen3.6-27B-Q4_K_M	Q4_K_M	NVIDIA GB10	32,768	78.4	2138.1	4717.7	375.03	973.18	56	1.41
Qwen3.6-35B-A3B-APEX-I-Balanced	Balanced	NVIDIA GeForce RTX 5090	32,768	78.1	247.3	630.1	11.87	15.28	150	0.52
Qwen3.6-35B-A3B-UD-Q5_K_M	Q5_K_M	Apple M4 Pro	8,192	77.5	831.1	1774.2	199.37	437.98	36	2.14
Qwen3.6-35B-A3B-UD-Q5_K_XL	Q5_K_XL	Apple M4 Pro	32,768	76.9	1650.3	5907.9	397.05	917.48	35	2.19
Qwen3.6-35B-A3B-UD-Q5_K_S	Q5_K_S	Apple M4 Pro	16,384	75.6	2124.5	9159.2	405.05	952.74	34	2.22
Qwen3.6-35B-A3B-UD-Q8_K_XL	Q8_K_XL	Apple M4 Pro	130,064	73.5	2541.3	10081.5	410.74	993.91	—	—
Qwen3.5-4B-BF16	BF16	NVIDIA GB10	130,064	72.5	304.9	557.5	53.63	57.62	44	1.65
Qwen3.6-27B-Q8_0	Q8_0	NVIDIA GeForce RTX 5090	8,192	72.2	1879.8	3967.1	198.94	544.61	139	0.52
Qwen3.6-27B-Q5_K_M	Q5_K_M	NVIDIA GB10	32,768	71.2	2497.3	8374.4	410.53	978.79	54	1.32
Qwen3.5-4B-Q8_0	Q8_0	Apple M4 Pro	8,192	70.8	91662.9	102865.1	55.72	97.24	38	1.88
Qwen3.5-4B-UD-Q8_K_XL	Q8_K_XL	Apple M4 Pro	8,192	68.2	40647.7	46032.2	57.67	110.83	38	1.80
Qwen3.5-4B-UD-IQ2_XXS	IQ2_XXS	Apple M4 Pro	130,064	67.2	415.0	1735.6	58.13	59.74	34	1.96
Qwen3.5-4B-Q4_1	Q4_1	Apple M4 Pro	130,064	66.5	42444.6	47515.9	59.31	94.11	35	1.93
Qwen3.5-4B-Q4_0	Q4_0	Apple M4 Pro	8,192	66.3	42796.3	47621.7	59.47	92.92	35	1.92
Qwen3.6-35B-A3B-Q8_0	Q8_0	NVIDIA GeForce RTX 4060 Ti	8,192	64.7	1345.0	4905.5	232.17	587.06	42	1.56
Qwen3.5-9B-BF16	BF16	NVIDIA GeForce RTX 5060 Ti	8,192	64.7	1319.1	3324.4	237.37	752.84	50	1.30
Qwen3.5-4B-UD-IQ2_M	IQ2_M	Apple M4 Pro	16,384	63.9	101118.4	114129.9	61.98	102.93	36	1.79
Qwen3.5-4B-UD-IQ3_XXS	IQ3_XXS	Apple M4 Pro	8,192	63.2	102853.0	115371.0	62.62	102.17	36	1.77
Qwen3.5-4B-Q4_K_S	Q4_K_S	Apple M4 Pro	32,768	62.1	269.9	982.9	30.86	38.52	37	1.67
Qwen3.5-4B-IQ4_XS	IQ4_XS	Apple M4 Pro	130,064	61.2	298.8	691.6	63.53	99.88	36	1.71
Qwen3.5-4B-UD-Q2_K_XL	Q2_K_XL	Apple M4 Pro	130,064	60.8	426.8	1782.3	64.36	65.87	33	1.84
Qwen3.6-35B-A3B-UD-Q2_K_XL	Q2_K_XL	NVIDIA GeForce RTX 5060 Ti	32,768	60.6	247.0	642.1	15.57	18.04	69	0.88
Qwen3.5-4B-UD-Q3_K_XL	Q3_K_XL	Apple M4 Pro	130,064	60.3	105937.7	120731.8	65.56	104.19	35	1.75
Qwen3.5-4B-Q3_K_M	Q3_K_M	Apple M4 Pro	32,768	60.2	107140.8	121084.9	65.76	107.02	34	1.80
Qwen3.5-4B-Q4_K_M	Q4_K_M	Apple M4 Pro	130,064	60.1	109129.1	121271.3	65.85	100.23	34	1.77
Qwen3.5-4B-UD-Q4_K_XL	Q4_K_XL	Apple M4 Pro	16,384	59.8	111004.3	121887.0	66.19	105.68	34	1.76
Qwen3.5-4B-Q3_K_S	Q3_K_S	Apple M4 Pro	16,384	58.7	111911.0	124051.4	67.30	117.02	33	1.76
Qwen3.5-4B-UD-Q6_K_XL	Q6_K_XL	Apple M4 Pro	32,768	58.2	111881.1	125038.3	68.00	110.23	36	1.62
Qwen3.6-35B-A3B-UD-IQ3_XXS	IQ3_XXS	NVIDIA GeForce RTX 5060 Ti	32,768	58.2	245.8	650.4	16.21	18.09	75	0.78
Qwen3.5-4B-Q5_K_S	Q5_K_S	Apple M4 Pro	130,064	57.8	113006.4	126019.9	68.32	122.07	34	1.69
DeepSeek-R1-Distill-Qwen-32B-Q8_0	Q8_0	NVIDIA GeForce RTX 5090	8,192	56.5	1640.1	3964.0	251.23	626.67	119	0.48
Qwen3.5-4B-Q5_K_M	Q5_K_M	Apple M4 Pro	130,064	56.4	115470.8	129146.4	70.16	119.78	34	1.65
Qwen3.5-4B-UD-Q5_K_XL	Q5_K_XL	Apple M4 Pro	130,064	56.3	50872.8	56228.1	70.35	113.69	34	1.66
Qwen3.6-35B-A3B-UD-IQ3_S	IQ3_S	NVIDIA GeForce RTX 5060 Ti	16,384	55.8	265.7	682.0	16.95	19.13	74	0.75
Qwen3.5-27B-UD-Q6_K_XL	Q6_K_XL	NVIDIA GeForce RTX 5090	8,192	55.3	2606.2	4481.2	542.39	1517.24	157	0.35
Qwen3.6-35B-A3B-UD-Q8_K_XL	Q8_K_XL	NVIDIA GeForce RTX 4060 Ti	8,192	54.9	1703.9	6414.9	276.14	591.31	39	1.39
Qwen3.6-27B-Q5_K_S	Q5_K_S	NVIDIA GB10	8,192	54.4	1674.5	3510.2	259.15	574.89	53	1.02
Qwen3.5-4B-Q6_K	Q6_K	Apple M4 Pro	16,384	53.7	298.9	1022.5	36.23	38.60	38	1.42
Qwen3.5-4B-IQ4_NL	IQ4_NL	Apple M4 Pro	8,192	50.5	424.5	1612.6	76.93	79.88	34	1.48
Qwen3.5-27B-UD-IQ2_XXS	IQ2_XXS	NVIDIA GB10	65,536	50.5	3412.5	16283.0	614.16	840.26	59	0.86
Qwen3.5-27B-IQ4_NL	IQ4_NL	NVIDIA GB10	130,064	48.8	2468.5	5767.8	624.14	1203.60	59	0.82
Qwen3.5-27B-Q4_0	Q4_0	NVIDIA GB10	32,768	48.6	2665.5	8717.5	629.31	1157.56	59	0.82
Qwen3.5-27B-IQ4_XS	IQ4_XS	NVIDIA GB10	32,768	48.6	3167.8	11401.0	626.64	1095.45	59	0.82
Qwen3.5-27B-Q3_K_S	Q3_K_S	NVIDIA GB10	65,536	48.1	2579.4	8242.8	634.82	1283.38	61	0.79
Qwen3.5-27B-Q3_K_M	Q3_K_M	NVIDIA GB10	32,768	47.9	2118.6	4828.6	630.30	1289.45	59	0.81
Qwen3.5-27B.Q8_0	Q8_0	NVIDIA GeForce RTX 5090	8,192	47.2	2007.1	4736.8	330.05	438.24	146	0.32
Qwen3.5-27B-Q4_1	Q4_1	NVIDIA GB10	32,768	46.9	2605.2	9689.9	655.03	1183.81	56	0.83
Qwen3.5-27B-UD-IQ2_M	IQ2_M	NVIDIA GB10	130,064	46.8	3119.1	9485.0	651.32	1235.57	61	0.77
Qwen3.5-27B-Q4_K_S	Q4_K_S	NVIDIA GB10	130,064	46.0	1828.5	8177.7	664.45	1369.03	56	0.82
Qwen3.5-27B-UD-Q2_K_XL	Q2_K_XL	NVIDIA GB10	65,536	46.0	3259.2	14575.0	660.29	1064.35	59	0.79
Qwen3.5-27B-Q4_K_M	Q4_K_M	NVIDIA GB10	130,064	45.8	1907.6	4815.3	669.62	1300.67	57	0.80
Qwen3.5-4B-BF16	BF16	Apple M4 Pro	16,384	45.0	20415.0	25350.6	87.84	91.96	41	1.09
Qwen3.5-27B-UD-IQ3_XXS	IQ3_XXS	NVIDIA GB10	65,536	43.2	2535.3	7640.9	710.41	1298.77	58	0.74
Qwen3.5-27B-UD-Q3_K_XL	Q3_K_XL	NVIDIA GB10	65,536	42.0	2992.7	9410.5	723.99	1260.95	58	0.72
Qwen3.5-27B-Q5_K_S	Q5_K_S	NVIDIA GB10	32,768	41.6	2874.5	9020.0	730.72	1411.63	57	0.73
Qwen3.5-27B-Q5_K_M	Q5_K_M	NVIDIA GB10	32,768	40.5	3218.5	11798.4	751.70	1433.91	58	0.70
Qwen3.5-27B-UD-Q4_K_XL	Q4_K_XL	NVIDIA GB10	32,768	36.8	2869.8	9114.7	834.67	1491.65	52	0.71
Qwen3.5-27B-UD-Q5_K_XL	Q5_K_XL	NVIDIA GB10	65,536	35.9	2655.0	6776.1	851.26	1545.75	52	0.69
Qwen3.5-27B-Q6_K	Q6_K	NVIDIA GB10	65,536	35.3	2673.0	6067.8	868.06	1664.08	52	0.68
Qwen3.5-27B-UD-IQ2_XXS	IQ2_XXS	Apple M4 Pro	16,384	34.7	14236.9	80386.7	849.96	1764.87	34	1.03
Qwen3.5-27B-Q8_0	Q8_0	NVIDIA GB10	32,768	34.0	3047.7	10803.9	898.77	1723.05	49	0.69
Qwen3.5-27B-Q8_0	Q8_0	Apple M4 Pro	32,768	33.2	6858.4	30413.1	907.77	3148.99	36	0.93
Qwen3.5-27B-UD-IQ3_XXS	IQ3_XXS	Apple M4 Pro	32,768	32.2	10089.5	56781.7	922.32	2570.41	34	0.95
Qwen3.5-27B-Q4_1	Q4_1	Apple M4 Pro	16,384	32.1	10049.3	59686.7	931.06	2120.49	35	0.92
Qwen3.5-27B-Q3_K_M	Q3_K_M	Apple M4 Pro	32,768	32.0	6672.6	29383.9	936.93	3794.48	35	0.92
Qwen3.5-27B-UD-Q4_K_XL	Q4_K_XL	Apple M4 Pro	32,768	32.0	10363.3	56049.7	930.34	2724.96	33	0.96
Qwen3.5-27B-Q6_K	Q6_K	Apple M4 Pro	130,064	31.9	7812.6	39473.6	938.05	3710.58	35	0.90
Qwen3.5-27B-Q4_0	Q4_0	Apple M4 Pro	32,768	31.9	6838.4	34624.7	949.33	2978.34	35	0.91
Qwen3.5-27B-IQ4_NL	IQ4_NL	Apple M4 Pro	130,064	31.7	6917.9	39969.3	954.36	3147.05	34	0.93
Qwen3.5-27B-IQ4_XS	IQ4_XS	Apple M4 Pro	130,064	31.4	6686.3	30170.3	964.56	3146.97	35	0.89
Qwen3.5-27B-UD-Q5_K_XL	Q5_K_XL	Apple M4 Pro	16,384	31.3	8919.1	57197.1	948.79	3052.71	34	0.93
Qwen3.5-27B-Q3_K_S	Q3_K_S	Apple M4 Pro	32,768	31.3	6694.8	28410.0	955.81	3847.84	36	0.88
Qwen3.5-27B-UD-Q3_K_XL	Q3_K_XL	Apple M4 Pro	32,768	30.9	10181.6	52862.3	951.99	2685.96	34	0.91
Qwen3.5-27B-UD-IQ2_M	IQ2_M	Apple M4 Pro	16,384	30.9	8284.5	43437.7	972.29	2854.47	33	0.93
Qwen3.5-27B-Q4_K_M	Q4_K_M	Apple M4 Pro	130,064	30.7	9252.4	49827.6	976.33	2974.58	34	0.90
Qwen3.5-27B-Q4_K_S	Q4_K_S	Apple M4 Pro	16,384	30.4	8605.2	47639.8	989.40	2774.14	34	0.90
Qwen3.5-27B-UD-Q2_K_XL	Q2_K_XL	Apple M4 Pro	130,064	30.4	9541.8	51287.4	983.68	3277.40	34	0.90
Qwen3.5-27B-UD-Q8_K_XL	Q8_K_XL	Apple M4 Pro	16,384	30.3	9614.3	43726.1	973.79	3628.40	35	0.86
Qwen3.5-27B-UD-Q6_K_XL	Q6_K_XL	Apple M4 Pro	32,768	30.2	9154.7	53033.8	981.24	3109.00	35	0.87
Qwen3.5-27B-Q5_K_M	Q5_K_M	Apple M4 Pro	16,384	30.1	9683.3	49735.4	986.98	3356.98	34	0.88
Qwen3.5-27B-Q5_K_S	Q5_K_S	Apple M4 Pro	32,768	29.8	10936.4	57252.9	993.81	2773.56	33	0.91
Qwen3.5-27B-UD-Q6_K_XL	Q6_K_XL	NVIDIA GB10	130,064	29.1	3507.8	9507.1	1050.01	1841.51	49	0.59
Qwen3.5-27B-UD-Q8_K_XL	Q8_K_XL	NVIDIA GB10	65,536	28.1	3855.9	13394.5	1088.63	1923.82	47	0.60
Qwen3.5-27B-UD-Q8_K_XL	Q8_K_XL	NVIDIA GeForce RTX 5090	8,192	22.9	5271.4	10138.0	1302.32	3347.34	105	0.22

791 of 791 entries

Deduplication policy: For each model + quantization + GPU combination, the run with the highest throughput (tok/s) is shown. Repeated runs are preserved in the raw dataset for further analysis.