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

143 open-weight models in creative writing 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	1460	7K	753B	Custom	2.1M	2,395
2	GLM-5.2zai-org	1459	11.1K	753B	MIT	318.5K	5,205
3	Kimi-K3moonshotai	1455	8.4K	2.8T	Custom	771.6K	11,950
4	GLM-5.1zai-org	1453	13.4K	754B	MIT	892.7K	1,850
5	DeepSeek-V4-Prodeepseek-ai	1446	9.7K	1.6T	MIT	335.6K	5,699
6	GLM-5.3-Flashzai-org	1444	10.6K	321B	MIT	6.4M	3,088
7	GLM-5zai-org	1438	4.9K	754B	MIT	1.4M	2,141
8	MiMo-V2.5-ProXiaomiMiMo	1437	14.6K	1.0T	MIT	9.9K	771
9	Kimi-K2.6moonshotai	1432	6.3K	1.0T	Modified MIT	237.2K	1,641
10	Kimi-K2.5moonshotai	1422	12.2K	1.0T	Modified MIT	395K	2,901
11	gemma-4-31Bgoogle	1415	971	31.3B	Apache 2.0	57K	589
12	GLM-4.6zai-org	1412	5.1K	357B	MIT	28.1K	1,247
13	Hy3tencent	1406	3.1K	299B	Apache 2.0	1.2M	990
14	DeepSeek-R1-0528deepseek-ai	1405	2.4K	685B	MIT	122.3K	2,495
15	DeepSeek-V3.2-Expdeepseek-ai	1404	1.7K	685B	MIT	43.6K	1,027
16	DeepSeek-V3.1-Terminusdeepseek-ai	1403	446	685B	MIT	9.5K	386
17	DeepSeek-V4-Flashdeepseek-ai	1403	8.5K	291B	MIT	1.1M	2,378
18	Qwen3.5-397B-A17BQwen	1402	17.8K	403B	Apache 2.0	857.2K	1,581
19	GLM-4.7zai-org	1401	2K	358B	MIT	66.6K	2,077
20	gemma-4-26B-A4Bgoogle	1401	1K	25.8B	Apache 2.0	74.1K	465
21	DeepSeek-V3.1deepseek-ai	1401	1.6K	685B	MIT	262.8K	864
22	DeepSeek-V3.2deepseek-ai	1399	6.9K	685B	MIT	7.5M	1,596
23	NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16nvidia	1398	2.2K	561B	OpenMDW 1.1	1.1M	384
24	MiniMax-M3MiniMaxAI	1396	13.6K	427B	Custom	155.9K	1,600
25	GLM-4.5zai-org	1395	3.1K	358B	MIT	110.4K	1,439
26	Inklingthinkingmachines	1395	11K	952B	Apache 2.0	1.8M	1,842
27	MiMo-V2.5XiaomiMiMo	1392	7.9K	311B	MIT	110.3K	434
28	Beam-1.0-501B-A23Breflection	1386	973	501B	Apache 2.0	11.8K	311
29	Mistral-Large-3-675B-Instruct-2512mistralai	1386	16.4K	—	Apache 2.0	2.5K	261
30	Qwen3-235B-A22B-Thinking-2507Qwen	1384	1.1K	235B	Apache 2.0	8.9K	419
31	NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-BF16nvidia	1377	975	124B	OpenMDW 1.1	70.3K	116
32	Qwen3.8-27BQwen	1377	6.6K	27.8B	Apache 2.0	7.8M	19,348
33	MiMo-V2-FlashXiaomiMiMo	1376	7.3K	310B	MIT	22.2K	756
34	Qwen3-235B-A22B-Instruct-2507Qwen	1375	13.9K	235B	Apache 2.0	84.8K	814
35	Mistral-Medium-3.5-128Bmistralai	1374	2.1K	128B	Custom	17.2K	498
36	Qwen3.5-122B-A10BQwen	1368	4.5K	125B	Apache 2.0	924.6K	653
37	Qwen3-VL-235B-A22B-InstructQwen	1366	1.4K	236B	Apache 2.0	327K	431
38	DeepSeek-V3-0324deepseek-ai	1364	6K	685B	MIT	153.8K	3,223
39	Qwen3.5-27BQwen	1363	4.3K	27.8B	Apache 2.0	1.7M	1,090
40	Step-3.5-Flashstepfun-ai	1357	9.7K	199B	Apache 2.0	219K	847
41	Qwen3-235B-A22BQwen	1356	4.6K	235B	Apache 2.0	308.9K	1,128
42	MiniMax-M2.7MiniMaxAI	1354	15.7K	229B	Custom	307.1K	1,261
43	DeepSeek-R1deepseek-ai	1354	3.3K	684B	MIT	1.6M	14,439
44	Qwen3.5-35B-A3BQwen	1348	4.7K	36.0B	Apache 2.0	1.6M	1,555
45	GLM-4.6Vzai-org	1347	425	108B	MIT	5.5K	398
46	LongCat-Flash-Chatmeituan-longcat	1346	1.3K	562B	MIT	13.3K	540
47	gemma-3-27b-itgoogle	1345	6.3K	27.4B	Gemma	364.5K	2,066
48	Qwen3-VL-235B-A22B-ThinkingQwen	1344	1K	236B	Apache 2.0	81K	420
49	GLM-4.5-Airzai-org	1343	4K	110B	MIT	1.3M	649
50	Hy3-previewtencent	1338	1K	299B	Custom	68.3K	314
51	Inkling-Smallthinkingmachines	1338	8.4K	266B	Apache 2.0	1.2M	449
52	Muse-Glimmer-30Bmeta-models	1337	725	29.8B	Apache 2.0	187.5K	2,127
53	Qwen3-Next-80B-A3B-InstructQwen	1333	3K	81.3B	Apache 2.0	129.9K	1,072
54	Qwen3-Coder-480B-A35B-InstructQwen	1333	3.2K	480B	Apache 2.0	58.6K	1,387
55	gemma-3-12b-itgoogle	1331	600	12.2B	Gemma	365.8K	908
56	MiniMax-M2.5MiniMaxAI	1330	6.7K	229B	Modified MIT	1.2M	1,529
57	DeepSeek-V3deepseek-ai	1329	3.6K	684B	—	1.6M	4,462
58	Trinity-Large-Previewarcee-ai	1319	4.8K	399B	OpenMDW 1.1	1.7K	181
59	Qwen3-30B-A3B-Instruct-2507Qwen	1317	3K	30.5B	Apache 2.0	932.4K	855
60	NVIDIA-Nemotron-3-Super-120B-A12B-BF16nvidia	1316	1.1K	124B	Custom	1.1M	438
61	INTELLECT-3PrimeIntellect	1316	833	107B	MIT	774	218
62	Qwen3-Next-80B-A3B-ThinkingQwen	1315	1.8K	81.3B	Apache 2.0	44.4K	503
63	Trinity-Large-Thinkingarcee-ai	1308	4.7K	399B	OpenMDW 1.1	3.2K	190
64	Mistral-Small-3.2-24B-Instruct-2506mistralai	1304	2.1K	24.0B	Apache 2.0	109.2K	644
65	Qwen3-32BQwen	1297	616	32.8B	Apache 2.0	3M	791
66	GLM-4.7-Flashzai-org	1295	1.9K	31.2B	MIT	1M	1,896
67	GLM-4.5Vzai-org	1295	613	108B	MIT	16.2K	726
68	QwQ-32BQwen	1287	3.7K	32.8B	Apache 2.0	32K	2,971
69	Ling-flash-2.0inclusionAI	1287	851	103B	MIT	4.4K	222
70	gemma-3n-E4B-itgoogle	1287	2.9K	7.8B	Gemma	12.7K	938
71	MiniMax-M2MiniMaxAI	1287	894	229B	Modified MIT	109.4K	1,514
72	granite-4.2-30bibm-granite	1285	656	29.3B	Apache 2.0	32.9K	140
73	DeepSeek-V2.5-1210deepseek-ai	1285	1.1K	236B	Custom	2.1K	275
74	gpt-oss-120bopenai	1275	3.9K	117B	Apache 2.0	2.4M	5,498
75	Qwen3-30B-A3BQwen	1271	3.3K	30.5B	Apache 2.0	1.1M	982
76	gemma-3-4b-itgoogle	1270	666	4.3B	Gemma	1M	1,633
77	Ring-flash-2.0inclusionAI	1270	888	103B	MIT	1.4K	104
78	NVIDIA-Nemotron-3-Nano-30B-A3B-BF16nvidia	1270	2.4K	31.6B	Custom	1.6M	840
79	Llama-3.1-Nemotron-70B-Instructnvidia	1268	1K	—	Llama 3.1	74	570
80	Llama-4-Maverick-17B-128E-Instructmeta-llama	1266	5.2K	402B	Llama 4	11.7K	520
81	Olmo-3.1-32B-Instructallenai	1265	1.7K	32.2B	Apache 2.0	20.8K	88
82	Llama-3.1-405B-Instructmeta-llama	1262	8.9K	406B	Llama 3.1	9.2K	607
83	NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16nvidia	1261	5.7K	31.6B	OpenMDW 1.1	1.1M	271
84	Olmo-3-32B-Thinkallenai	1255	860	32.2B	Apache 2.0	12.2K	177
85	Mistral-Small-3.1-24B-Instruct-2503mistralai	1252	4.1K	24.0B	Apache 2.0	168.4K	1,408
86	Llama-3.3-70B-Instructmeta-llama	1250	8K	70.6B	Llama 3.3	357.4K	3,197
87	Llama-4-Scout-17B-16E-Instructmeta-llama	1249	3.8K	109B	Llama 4	110.7K	1,387
88	granite-4.1-8bibm-granite	1249	692	8.8B	Apache 2.0	98.1K	257
89	gemma-2-27b-itgoogle	1240	11.4K	27.2B	Gemma	21.7K	577
90	DeepSeek-V2.5deepseek-ai	1236	3.5K	236B	Custom	4.5K	750
91	granite-4.2-8bibm-granite	1236	639	8.8B	Apache 2.0	69.2K	116
92	Llama-3.1-70B-Instructmeta-llama	1232	8.3K	70.6B	Llama 3.1	92K	1,027
93	Llama-3.1-Tulu-3-70Ballenai	1231	441	70.6B	Llama 3.1	965	61
94	Olmo-3.1-32B-Thinkallenai	1226	1.4K	32.2B	Apache 2.0	7K	120
95	Qwen2.5-72B-InstructQwen	1221	5.7K	72.7B	Custom	192.8K	1,043
96	granite-4.0-h-smallibm-granite	1210	706	32.2B	Apache 2.0	23.5K	313
97	Meta-Llama-3-70B-Instructmeta-llama	1209	22.8K	70.6B	Llama 3	37.8K	1,537
98	granite-4.2-3bibm-granite	1206	632	3.7B	Apache 2.0	38.9K	131
99	gemma-2-9b-itgoogle	1205	8.3K	9.2B	Gemma	113.6K	1,166
100	Nemotron-4-340B-Instructnvidia	1202	3.2K	—	Custom	638	703
101	gpt-oss-20bopenai	1201	1.2K	20.9B	Apache 2.0	5.9M	5,286
102	aya-expanse-32bCohereLabs	1199	3.9K	32.3B	CC BY-NC 4.0	3K	303
103	Mistral-Small-24B-Instruct-2501mistralai	1194	2.4K	23.6B	Apache 2.0	94.4K	991
104	phi-4microsoft	1181	4.1K	14.7B	MIT	293.7K	2,340
105	Llama-3.1-Tulu-3-8Ballenai	1181	472	8.0B	Llama 3.1	5K	178
106	Qwen2-72B-InstructQwen	1181	5.5K	72.7B	Tongyi Qianwen	24.3K	724
107	Qwen2.5-Coder-32B-InstructQwen	1174	836	32.8B	Apache 2.0	548.8K	2,206
108	aya-expanse-8bCohereLabs	1168	1.5K	8.0B	CC BY-NC 4.0	22.9K	463
109	Llama-3.1-8B-Instructmeta-llama	1154	7.3K	8.0B	Llama 3.1	4.3M	8,836
110	Meta-Llama-3-8B-Instructmeta-llama	1149	15.4K	8.0B	Llama 3	359.6K	5,787
111	Qwen1.5-110B-ChatQwen	1148	3.7K	111B	Tongyi Qianwen	2.7K	139
112	gemma-2-2b-itgoogle	1146	7.1K	2.6B	Gemma	125.7K	1,607
113	Mixtral-8x22B-Instruct-v0.1mistralai	1141	7.4K	141B	Apache 2.0	38.7K	780
114	Qwen1.5-72B-ChatQwen	1137	5.9K	72.3B	Tongyi Qianwen	11.2K	224
115	QwQ-32B-PreviewQwen	1133	454	32.8B	Apache 2.0	7.1K	1,747
116	granite-3.1-8b-instructibm-granite	1129	450	8.2B	Apache 2.0	27.4K	169
117	DeepSeek-Coder-V2-Instructdeepseek-ai	1120	2.3K	236B	Custom	3.8K	718
118	granite-3.1-2b-instructibm-granite	1116	442	2.5B	Apache 2.0	24.5K	57
119	Mixtral-8x7B-Instruct-v0.1mistralai	1108	11.2K	46.7B	Apache 2.0	298.7K	4,777
120	Phi-3-medium-4k-instructmicrosoft	1106	3.7K	14.0B	MIT	18.7K	231
121	Llama-3.2-3B-Instructmeta-llama	1094	1.1K	3.2B	Llama 3.2	914.1K	2,963
122	Qwen1.5-14B-ChatQwen	1090	2.4K	14.2B	Tongyi Qianwen	12.2K	118
123	SOLAR-10.7B-Instruct-v1.0upstage	1089	698	10.7B	CC BY-NC 4.0	48.7K	660
124	Qwen1.5-32B-ChatQwen	1083	3K	32.5B	Tongyi Qianwen	46.8K	115
125	Phi-3-small-8k-instructmicrosoft	1082	2.4K	7.4B	MIT	7.1K	182
126	Llama-2-70b-chat-hfmeta-llama	1075	6.4K	69.0B	Llama 2	4.8K	2,211
127	granite-3.0-8b-instructibm-granite	1070	895	8.2B	Apache 2.0	13.3K	209
128	Mistral-7B-Instruct-v0.2mistralai	1067	3K	7.2B	Apache 2.0	1.6M	3,280
129	deepseek-llm-67b-chatdeepseek-ai	1067	789	—	Custom	2.8K	222
130	gemma-1.1-7b-itgoogle	1059	3.4K	8.5B	Gemma	4.2K	276
131	Llama-2-13b-chat-hfmeta-llama	1047	3.3K	13.0B	Llama 2	148.8K	1,138
132	granite-3.0-2b-instructibm-granite	1046	916	2.6B	Apache 2.0	3.2K	48
133	Phi-3-mini-4k-instructmicrosoft	1044	2K	3.8B	MIT	329.2K	1,489
134	Qwen1.5-7B-ChatQwen	1034	701	7.7B	Tongyi Qianwen	7K	192
135	Llama-2-7b-chat-hfmeta-llama	1033	2.3K	6.7B	Llama 2	126.4K	4,951
136	Llama-3.2-1B-Instructmeta-llama	1032	1.1K	1.2B	Llama 3.2	1.5M	1,961
137	Qwen-14B-ChatQwen	1027	898	14.2B	—	5.2K	381
138	Phi-3-mini-128k-instructmicrosoft	1024	2.9K	3.8B	MIT	238.7K	1,713
139	gemma-7b-itgoogle	1023	1.3K	8.5B	Gemma	25.3K	1,263
140	gemma-1.1-2b-itgoogle	998	1.5K	2.5B	Gemma	6.1K	175
141	OLMo-7B-Instructallenai	989	1K	6.9B	Apache 2.0	846	54
142	gemma-2b-itgoogle	986	696	2.5B	Gemma	14K	975
143	Qwen1.5-4B-ChatQwen	969	1.1K	4.0B	Custom	15.9K	53

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.