Ran it, or it didn’t happen.

Benchmark open models on your own hardware. Runs count once their event log checks out.

Recent verified runsWhereSpeed

PassedFailedNot supported

Cloud leaderboard

Every model runs on the same pinned machine. 17 models, 70 runs.

Suite: [email protected]
Machine: NVIDIA L40S · 48 GB
Engine: ollama 0.40.14

#	Model	Pass rate				Speed	Runs
1	Qwen3.5-35B-A3BQ4_K_M	96%	9/9	8/9	9/9	201.4 tok/s	3
2	Qwen3.5-27BQ4_K_M	93%	8/9	9/9	8/9	35.5 tok/s	3
3	Qwen3-30B-A3B-Instruct-2507Q4_K_M	92%	11/12	10/12	12/12	187.6 tok/s	4
4	GLM-4.7-FlashQ4_K_M	89%	9/9	6/9	9/9	201.7 tok/s	3
5	Qwen3-Coder-30B-A3B-InstructQ4_K_M	89%	9/9	7/9	8/9	186.4 tok/s	3
6	Qwen3-8BQ4_K_M	89%	9/12	11/12	12/12	120.7 tok/s	4
7	Qwen3-14BQ4_K_M	87%	13/15	11/15	15/15	67.1 tok/s	5
8	gemma-3-27b-itQ4_K_M	86%	17/18	Not supported	14/18	36.2 tok/s	6
9	gpt-oss-20bMXFP4	85%	9/9	9/9	5/9	173.8 tok/s	3
10	gemma-3-4b-itQ4_K_M	83%	11/15	Not supported	14/15	232.7 tok/s	5
11	Qwen3.5-2BQ4_K_M	81%	9/12	9/12	11/12	298.1 tok/s	4
12	Qwen3-4BQ4_K_M	78%	6/9	6/9	9/9	247.9 tok/s	3
13	gemma-3-12b-itQ4_K_M	75%	8/12	Not supported	10/12	80.9 tok/s	4
14	Llama-3.1-8B-InstructQ4_K_M	71%	8/15	9/15	15/15	124.4 tok/s	5
15	Phi-3.5-mini-instructQ4_K_M	64%	9/18	Not supported	14/18	261.3 tok/s	6
16	MiniCPM5-2BQ4_K_M	52%	4/9	3/9	7/9	301.4 tok/s	3
17	Mistral-7B-Instruct-v0.2Q4_K_M	47%	6/18	Not supported	11/18	137.1 tok/s	6

Ranked by checks passed. A task a model can’t do is shown, not counted as a zero.

Open model leaderboard

135 open-weight models in expert ranked by LMArena score · updated Nov 6, 2026

Top rated

Most downloaded

Newest

All sizes ≤3B 3–9B 9–20B 20–40B 40–80B >80B

#	Model	Score	Votes	Params	License	Downloads	Likes
1	GLM-5.3zai-org	1519	3K	753B	Custom	2.1M	2,395
2	Kimi-K3moonshotai	1518	3.7K	2.8T	Custom	771.6K	11,950
3	GLM-5.3-Flashzai-org	1511	5.1K	321B	MIT	6.4M	3,088
4	MiMo-V2.5-ProXiaomiMiMo	1501	7.7K	1.0T	MIT	9.9K	771
5	Kimi-K2.6moonshotai	1491	3.9K	1.0T	Modified MIT	237.2K	1,641
6	GLM-5.2zai-org	1489	5.7K	753B	MIT	318.5K	5,205
7	Qwen3.8-27BQwen	1479	3.5K	27.8B	Apache 2.0	7.8M	19,348
8	GLM-5.1zai-org	1478	7.3K	754B	MIT	892.7K	1,850
9	NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16nvidia	1471	1.2K	561B	OpenMDW 1.1	1.1M	384
10	Inklingthinkingmachines	1469	5.8K	952B	Apache 2.0	1.8M	1,842
11	Kimi-K2.5moonshotai	1466	7K	1.0T	Modified MIT	395K	2,901
12	gemma-4-31Bgoogle	1465	502	31.3B	Apache 2.0	57K	589
13	Qwen3-235B-A22B-Thinking-2507Qwen	1463	423	235B	Apache 2.0	8.9K	419
14	DeepSeek-V4-Prodeepseek-ai	1463	5.9K	1.6T	MIT	335.6K	5,699
15	Qwen3.5-397B-A17BQwen	1462	10.7K	403B	Apache 2.0	857.2K	1,581
16	MiMo-V2.5XiaomiMiMo	1460	4.8K	311B	MIT	110.3K	434
17	MiniMax-M3MiniMaxAI	1459	7.9K	427B	Custom	155.9K	1,600
18	Hy3tencent	1458	1.6K	299B	Apache 2.0	1.2M	990
19	Beam-1.0-501B-A23Breflection	1455	413	501B	Apache 2.0	11.8K	311
20	GLM-5zai-org	1454	2.6K	754B	MIT	1.4M	2,141
21	gemma-4-26B-A4Bgoogle	1447	453	25.8B	Apache 2.0	74.1K	465
22	LongCat-Flash-Chatmeituan-longcat	1444	519	562B	MIT	13.3K	540
23	MiniMax-M2.7MiniMaxAI	1443	9.1K	229B	Custom	307.1K	1,261
24	Qwen3-VL-235B-A22B-InstructQwen	1442	595	236B	Apache 2.0	327K	431
25	Qwen3-235B-A22B-Instruct-2507Qwen	1442	6.4K	235B	Apache 2.0	84.8K	814
26	Inkling-Smallthinkingmachines	1441	4.5K	266B	Apache 2.0	1.2M	449
27	DeepSeek-V4-Flashdeepseek-ai	1441	5.4K	291B	MIT	1.1M	2,378
28	DeepSeek-V3.2deepseek-ai	1436	3.4K	685B	MIT	7.5M	1,596
29	GLM-4.5zai-org	1433	1.2K	358B	MIT	110.4K	1,439
30	Mistral-Medium-3.5-128Bmistralai	1432	1.2K	128B	Custom	17.2K	498
31	GLM-4.6zai-org	1432	2.1K	357B	MIT	28.1K	1,247
32	Qwen3.5-122B-A10BQwen	1432	2.7K	125B	Apache 2.0	924.6K	653
33	Qwen3-VL-235B-A22B-ThinkingQwen	1429	412	236B	Apache 2.0	81K	420
34	Qwen3.5-27BQwen	1428	2.6K	27.8B	Apache 2.0	1.7M	1,090
35	NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-BF16nvidia	1427	467	124B	OpenMDW 1.1	70.3K	116
36	MiMo-V2-FlashXiaomiMiMo	1425	4K	310B	MIT	22.2K	756
37	GLM-4.7zai-org	1423	808	358B	MIT	66.6K	2,077
38	Hy3-previewtencent	1423	707	299B	Custom	68.3K	314
39	DeepSeek-V3.2-Expdeepseek-ai	1422	424	685B	MIT	43.6K	1,027
40	Step-3.5-Flashstepfun-ai	1421	5.3K	199B	Apache 2.0	219K	847
41	Mistral-Large-3-675B-Instruct-2512mistralai	1421	8.7K	—	Apache 2.0	2.5K	261
42	Qwen3-Next-80B-A3B-InstructQwen	1416	1.1K	81.3B	Apache 2.0	129.9K	1,072
43	Qwen3.5-35B-A3BQwen	1407	2.8K	36.0B	Apache 2.0	1.6M	1,555
44	granite-4.2-30bibm-granite	1407	365	29.3B	Apache 2.0	32.9K	140
45	DeepSeek-V3.1deepseek-ai	1405	538	685B	MIT	262.8K	864
46	NVIDIA-Nemotron-3-Super-120B-A12B-BF16nvidia	1398	657	124B	Custom	1.1M	438
47	Qwen3-30B-A3B-Instruct-2507Qwen	1397	1.2K	30.5B	Apache 2.0	932.4K	855
48	DeepSeek-R1-0528deepseek-ai	1394	886	685B	MIT	122.3K	2,495
49	Muse-Glimmer-30Bmeta-models	1388	457	29.8B	Apache 2.0	187.5K	2,127
50	granite-4.2-8bibm-granite	1381	313	8.8B	Apache 2.0	69.2K	116
51	MiniMax-M2.5MiniMaxAI	1378	3.8K	229B	Modified MIT	1.2M	1,529
52	Qwen3-Next-80B-A3B-ThinkingQwen	1375	662	81.3B	Apache 2.0	44.4K	503
53	GLM-4.6Vzai-org	1373	185	108B	MIT	5.5K	398
54	GLM-4.5-Airzai-org	1370	1.5K	110B	MIT	1.3M	649
55	Qwen3-235B-A22BQwen	1369	1.9K	235B	Apache 2.0	308.9K	1,128
56	Qwen3-32BQwen	1362	236	32.8B	Apache 2.0	3M	791
57	NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16nvidia	1361	2.8K	31.6B	OpenMDW 1.1	1.1M	271
58	Trinity-Large-Thinkingarcee-ai	1360	2.9K	399B	OpenMDW 1.1	3.2K	190
59	GLM-4.7-Flashzai-org	1357	891	31.2B	MIT	1M	1,896
60	gpt-oss-120bopenai	1356	1.4K	117B	Apache 2.0	2.4M	5,498
61	NVIDIA-Nemotron-3-Nano-30B-A3B-BF16nvidia	1356	1K	31.6B	Custom	1.6M	840
62	Ling-flash-2.0inclusionAI	1355	353	103B	MIT	4.4K	222
63	GLM-4.5Vzai-org	1353	190	108B	MIT	16.2K	726
64	Trinity-Large-Previewarcee-ai	1351	2.9K	399B	OpenMDW 1.1	1.7K	181
65	DeepSeek-V3-0324deepseek-ai	1350	2.4K	685B	MIT	153.8K	3,223
66	Ring-flash-2.0inclusionAI	1344	358	103B	MIT	1.4K	104
67	MiniMax-M2MiniMaxAI	1339	300	229B	Modified MIT	109.4K	1,514
68	Qwen3-Coder-480B-A35B-InstructQwen	1338	1.4K	480B	Apache 2.0	58.6K	1,387
69	DeepSeek-R1deepseek-ai	1337	848	684B	MIT	1.6M	14,439
70	QwQ-32BQwen	1323	1.2K	32.8B	Apache 2.0	32K	2,971
71	INTELLECT-3PrimeIntellect	1322	354	107B	MIT	774	218
72	Qwen3-30B-A3BQwen	1314	1.3K	30.5B	Apache 2.0	1.1M	982
73	granite-4.1-8bibm-granite	1309	408	8.8B	Apache 2.0	98.1K	257
74	granite-4.2-3bibm-granite	1309	312	3.7B	Apache 2.0	38.9K	131
75	Olmo-3.1-32B-Instructallenai	1308	761	32.2B	Apache 2.0	20.8K	88
76	DeepSeek-V3deepseek-ai	1305	1.2K	684B	—	1.6M	4,462
77	gemma-3-27b-itgoogle	1304	2.3K	27.4B	Gemma	364.5K	2,066
78	Olmo-3.1-32B-Thinkallenai	1295	552	32.2B	Apache 2.0	7K	120
79	Mistral-Small-3.2-24B-Instruct-2506mistralai	1291	856	24.0B	Apache 2.0	109.2K	644
80	Olmo-3-32B-Thinkallenai	1272	314	32.2B	Apache 2.0	12.2K	177
81	DeepSeek-V2.5-1210deepseek-ai	1266	441	236B	Custom	2.1K	275
82	Llama-4-Maverick-17B-128E-Instructmeta-llama	1259	2K	402B	Llama 4	11.7K	520
83	gpt-oss-20bopenai	1258	506	20.9B	Apache 2.0	5.9M	5,286
84	Mistral-Small-3.1-24B-Instruct-2503mistralai	1257	1.7K	24.0B	Apache 2.0	168.4K	1,408
85	granite-4.0-h-smallibm-granite	1251	316	32.2B	Apache 2.0	23.5K	313
86	gemma-3-12b-itgoogle	1248	186	12.2B	Gemma	365.8K	908
87	gemma-3n-E4B-itgoogle	1246	1.1K	7.8B	Gemma	12.7K	938
88	Qwen2.5-72B-InstructQwen	1245	2.4K	72.7B	Custom	192.8K	1,043
89	Llama-3.1-405B-Instructmeta-llama	1243	3.1K	406B	Llama 3.1	9.2K	607
90	Llama-3.1-Nemotron-70B-Instructnvidia	1242	453	—	Llama 3.1	74	570
91	DeepSeek-V2.5deepseek-ai	1240	1.5K	236B	Custom	4.5K	750
92	Llama-4-Scout-17B-16E-Instructmeta-llama	1235	1.5K	109B	Llama 4	110.7K	1,387
93	Llama-3.3-70B-Instructmeta-llama	1225	2.9K	70.6B	Llama 3.3	357.4K	3,197
94	gemma-3-4b-itgoogle	1223	208	4.3B	Gemma	1M	1,633
95	Qwen2.5-Coder-32B-InstructQwen	1220	267	32.8B	Apache 2.0	548.8K	2,206
96	Llama-3.1-70B-Instructmeta-llama	1209	2.9K	70.6B	Llama 3.1	92K	1,027
97	phi-4microsoft	1203	1.1K	14.7B	MIT	293.7K	2,340
98	Mistral-Small-24B-Instruct-2501mistralai	1202	754	23.6B	Apache 2.0	94.4K	991
99	aya-expanse-32bCohereLabs	1182	1.8K	32.3B	CC BY-NC 4.0	3K	303
100	DeepSeek-Coder-V2-Instructdeepseek-ai	1181	769	236B	Custom	3.8K	718
101	gemma-2-27b-itgoogle	1171	4K	27.2B	Gemma	21.7K	577
102	Nemotron-4-340B-Instructnvidia	1171	1K	—	Custom	638	703
103	Qwen2-72B-InstructQwen	1171	1.8K	72.7B	Tongyi Qianwen	24.3K	724
104	aya-expanse-8bCohereLabs	1152	600	8.0B	CC BY-NC 4.0	22.9K	463
105	Meta-Llama-3-70B-Instructmeta-llama	1149	8K	70.6B	Llama 3	37.8K	1,537
106	gemma-2-9b-itgoogle	1147	2.8K	9.2B	Gemma	113.6K	1,166
107	Qwen1.5-110B-ChatQwen	1144	1.4K	111B	Tongyi Qianwen	2.7K	139
108	Llama-3.1-8B-Instructmeta-llama	1144	2.6K	8.0B	Llama 3.1	4.3M	8,836
109	granite-3.1-8b-instructibm-granite	1141	237	8.2B	Apache 2.0	27.4K	169
110	Qwen1.5-72B-ChatQwen	1136	1.8K	72.3B	Tongyi Qianwen	11.2K	224
111	QwQ-32B-PreviewQwen	1134	232	32.8B	Apache 2.0	7.1K	1,747
112	granite-3.1-2b-instructibm-granite	1131	225	2.5B	Apache 2.0	24.5K	57
113	Qwen1.5-32B-ChatQwen	1126	1.2K	32.5B	Tongyi Qianwen	46.8K	115
114	Meta-Llama-3-8B-Instructmeta-llama	1113	5.4K	8.0B	Llama 3	359.6K	5,787
115	Mixtral-8x22B-Instruct-v0.1mistralai	1113	2.6K	141B	Apache 2.0	38.7K	780
116	Phi-3-medium-4k-instructmicrosoft	1108	1.1K	14.0B	MIT	18.7K	231
117	gemma-2-2b-itgoogle	1096	2.5K	2.6B	Gemma	125.7K	1,607
118	Qwen1.5-14B-ChatQwen	1094	944	14.2B	Tongyi Qianwen	12.2K	118
119	Llama-3.2-3B-Instructmeta-llama	1090	499	3.2B	Llama 3.2	914.1K	2,963
120	Mixtral-8x7B-Instruct-v0.1mistralai	1088	3.2K	46.7B	Apache 2.0	298.7K	4,777
121	granite-3.0-8b-instructibm-granite	1087	345	8.2B	Apache 2.0	13.3K	209
122	Phi-3-small-8k-instructmicrosoft	1067	895	7.4B	MIT	7.1K	182
123	granite-3.0-2b-instructibm-granite	1064	407	2.6B	Apache 2.0	3.2K	48
124	Qwen1.5-7B-ChatQwen	1055	231	7.7B	Tongyi Qianwen	7K	192
125	Phi-3-mini-4k-instructmicrosoft	1051	486	3.8B	MIT	329.2K	1,489
126	gemma-1.1-7b-itgoogle	1039	1.2K	8.5B	Gemma	4.2K	276
127	Llama-2-70b-chat-hfmeta-llama	1039	1.4K	69.0B	Llama 2	4.8K	2,211
128	Llama-2-7b-chat-hfmeta-llama	1036	411	6.7B	Llama 2	126.4K	4,951
129	Mistral-7B-Instruct-v0.2mistralai	1036	798	7.2B	Apache 2.0	1.6M	3,280
130	Llama-2-13b-chat-hfmeta-llama	1030	539	13.0B	Llama 2	148.8K	1,138
131	Llama-3.2-1B-Instructmeta-llama	1007	487	1.2B	Llama 3.2	1.5M	1,961
132	gemma-7b-itgoogle	1001	340	8.5B	Gemma	25.3K	1,263
133	Phi-3-mini-128k-instructmicrosoft	984	1.1K	3.8B	MIT	238.7K	1,713
134	Qwen1.5-4B-ChatQwen	980	356	4.0B	Custom	15.9K	53
135	gemma-1.1-2b-itgoogle	970	578	2.5B	Gemma	6.1K	175

Show the top 20

Scores: LMArena, CC BY 4.0. Measured on served versions, not local builds. A model with several served variants shows its best.