Agentic · tools

QwenClawBench leaderboard

Name: benchgap: LLM benchmark scores, measured and estimated
Creator: benchgap

As of 2026-10-07, the highest measured score on QwenClawBench is 64.3% by Qwen3.7 Max. 128 more models have estimated scores, calibrated from the benchmarks they were measured on.

Measured scores: benchlm.ai.

#	Model	Score	Source
1	Claude Fable 5	100.0%	estimated ± 6.0 pp, low confidence
2	Claude Fable 5.1	100.0%	estimated ± 6.0 pp, low confidence
3	Claude Opus 5.5	100.0%	estimated ± 6.0 pp, low confidence
4	Claude Sonnet 5	100.0%	estimated ± 6.0 pp, low confidence
5	Claude Sonnet 5.5	100.0%	estimated ± 6.0 pp, low confidence
6	DeepSeek V4.1 Flash	100.0%	estimated ± 6.0 pp, low confidence
7	Gemini 3.6 Flash	100.0%	estimated ± 6.0 pp, low confidence
8	Gemini 3.7 Flash	100.0%	estimated ± 6.0 pp, low confidence
9	Gemini 3.8 Flash	100.0%	estimated ± 6.0 pp, low confidence
10	Gemini 4 Argon	100.0%	estimated ± 6.0 pp, low confidence
11	GLM-5.3	100.0%	estimated ± 6.0 pp, low confidence
12	GPT-5.6 Luna	100.0%	estimated ± 6.0 pp, low confidence
13	GPT-5.6 Sol	100.0%	estimated ± 6.0 pp, low confidence
14	GPT-5.6 Terra	100.0%	estimated ± 6.0 pp, low confidence
15	GPT-6.1 Sol	100.0%	estimated ± 6.0 pp, low confidence
16	GPT-6 Astra	100.0%	estimated ± 6.0 pp, low confidence
17	GPT-6 Luna	100.0%	estimated ± 6.0 pp, low confidence
18	GPT-6 Sol	100.0%	estimated ± 6.0 pp, low confidence
19	Grok 4.5	100.0%	estimated ± 6.0 pp, low confidence
20	Grok 4.6	100.0%	estimated ± 6.0 pp, low confidence
21	Grok 4.7	100.0%	estimated ± 6.0 pp, low confidence
22	Hy3 Preview	100.0%	estimated ± 6.0 pp, low confidence
23	Ling 3.1 Flash	100.0%	estimated ± 6.0 pp, low confidence
24	MiMo-V2.6-Flash	100.0%	estimated ± 6.0 pp, low confidence
25	MiMo-V2.6-Pro	100.0%	estimated ± 6.0 pp, low confidence
26	Mistral Large 4	100.0%	estimated ± 6.0 pp, low confidence
27	Muse Spark 1.2	100.0%	estimated ± 6.0 pp, low confidence
28	Muse Spark 1.3	100.0%	estimated ± 6.0 pp, low confidence
29	Qwen3.8-27B	100.0%	estimated ± 6.0 pp, low confidence
30	Qwen3.8-Flash-Next	100.0%	estimated ± 6.0 pp, low confidence
31	Qwen3.8 Max Preview	100.0%	estimated ± 6.0 pp, low confidence
32	Apodex 1.1	99.4%	estimated ± 6.0 pp, low confidence
33	Apodex 1.1 Mini	99.4%	estimated ± 6.0 pp, low confidence
34	Quasar 438B	97.9%	estimated ± 6.0 pp, low confidence
35	Ling 3.0 Flash VL	89.8%	estimated ± 6.0 pp, low confidence
36	Nemotron 3 Ultra	88.5%	estimated ± 6.0 pp, low confidence
37	Muse Spark 1.1	73.0%	estimated ± 2.6 pp, medium confidence
38	Claude Opus 5	70.9%	estimated ± 2.6 pp, medium confidence
39	Step 5 Preview	70.7%	estimated ± 2.6 pp, medium confidence
40	Kimi K3	69.4%	estimated ± 2.6 pp, medium confidence
41	Hy4 preview	69.0%	estimated ± 2.6 pp, medium confidence
42	Gemini 3.5 Flash	68.9%	estimated ± 2.6 pp, medium confidence
43	Claude Opus 4.8	67.7%	estimated ± 2.6 pp, medium confidence
44	Ornith-1.5-397B	65.9%	estimated ± 2.6 pp, medium confidence
45	Inkling-Small	65.6%	estimated ± 2.6 pp, medium confidence
46	Beam	65.0%	estimated ± 2.6 pp, medium confidence
47	Qwen3.7 Max	64.3%	measured
48	Claude Opus 4.7 (Adaptive)	63.9%	estimated ± 2.6 pp, medium confidence
49	GLM-5.2	63.6%	estimated ± 2.6 pp, medium confidence
50	Kimi K2.7 Code	63.0%	estimated ± 2.6 pp, high confidence
51	Muse Glimmer 30B	62.7%	estimated ± 2.6 pp, high confidence
52	GPT-5.5	62.6%	estimated ± 2.6 pp, high confidence
53	MiniMax M3	61.9%	estimated ± 2.6 pp, high confidence
54	Inkling	61.8%	estimated ± 2.6 pp, high confidence
55	Qwen3.7 Plus	61.8%	measured
56	DeepSeek V4 Pro 0813	61.5%	estimated ± 2.6 pp, high confidence
57	MiMo-V2.5-Pro	60.7%	estimated ± 6.0 pp, low confidence
58	GLM-5.1	60.5%	estimated ± 2.6 pp, high confidence
59	GPT-5.4	59.8%	estimated ± 2.6 pp, high confidence
60	Ornith-1.5-35B-A3B	59.6%	estimated ± 2.6 pp, high confidence
61	DeepSeek V4 Flash 0731	59.1%	estimated ± 2.6 pp, high confidence
62	Qwen 3.6 Max (preview)	59.0%	measured
63	Ling 3.0 Flash	57.6%	estimated ± 2.6 pp, high confidence
64	Qwen3.6 Plus	57.2%	measured
65	Solar Pro 4	56.2%	estimated ± 2.6 pp, high confidence
66	Grok 4.3	56.1%	estimated ± 6.0 pp, low confidence
67	Hy3	55.9%	estimated ± 6.0 pp, low confidence
68	Step 3.7 Flash	55.9%	estimated ± 6.0 pp, low confidence
69	MiniMax M2.7	55.9%	estimated ± 6.0 pp, low confidence
70	Muse Spark	55.9%	estimated ± 6.0 pp, low confidence
71	A.X K2	55.9%	estimated ± 6.0 pp, low confidence
72	Celeris-1	55.9%	estimated ± 6.0 pp, low confidence
73	Command A+	55.9%	estimated ± 6.0 pp, low confidence
74	DeepSeek V3	55.9%	estimated ± 6.0 pp, low confidence
75	DeepSeek V3 0324	55.9%	estimated ± 6.0 pp, low confidence
76	Gemini 2.5 Pro	55.9%	estimated ± 6.0 pp, low confidence
77	Gemini 3.1 Pro	55.9%	estimated ± 6.0 pp, low confidence
78	Gemini 3.5 Flash-Lite	55.9%	estimated ± 6.0 pp, low confidence
79	Gemma 3 27B	55.9%	estimated ± 6.0 pp, low confidence
80	Gemma 4 12B	55.9%	estimated ± 6.0 pp, low confidence
81	Gemma 4 26B A4B	55.9%	estimated ± 6.0 pp, low confidence
82	Gemma 4 31B	55.9%	estimated ± 6.0 pp, low confidence
83	Gemma 4 E2B	55.9%	estimated ± 6.0 pp, low confidence
84	Gemma 4 E4B	55.9%	estimated ± 6.0 pp, low confidence
85	GPT-4.1 mini	55.9%	estimated ± 6.0 pp, low confidence
86	GPT-4.1 nano	55.9%	estimated ± 6.0 pp, low confidence
87	GPT-4o	55.9%	estimated ± 6.0 pp, low confidence
88	GPT-4o mini	55.9%	estimated ± 6.0 pp, low confidence
89	GPT-5.1	55.9%	estimated ± 6.0 pp, low confidence
90	GPT-5 (high)	55.9%	estimated ± 6.0 pp, low confidence
91	GPT-OSS 120B	55.9%	estimated ± 6.0 pp, low confidence
92	GPT-OSS 20B	55.9%	estimated ± 6.0 pp, low confidence
93	Granite 4.2 30B	55.9%	estimated ± 6.0 pp, low confidence
94	Granite 4.2 3B	55.9%	estimated ± 6.0 pp, low confidence
95	Granite 4.2 8B	55.9%	estimated ± 6.0 pp, low confidence
96	K-Exaone	55.9%	estimated ± 6.0 pp, low confidence
97	Kimi K2.5 (Reasoning)	55.9%	estimated ± 6.0 pp, low confidence
98	LFM2.5-2.6B	55.9%	estimated ± 6.0 pp, low confidence
99	Ling 2.6 Flash	55.9%	estimated ± 6.0 pp, low confidence
100	Ling 3.0 Flash FP8	55.9%	estimated ± 6.0 pp, low confidence
101	Ling 3.0 Tiny	55.9%	estimated ± 6.0 pp, low confidence
102	Llama 4 Maverick	55.9%	estimated ± 6.0 pp, low confidence
103	Llama 4 Scout	55.9%	estimated ± 6.0 pp, low confidence
104	Mercury 2.5	55.9%	estimated ± 6.0 pp, low confidence
105	MiMo-V2-Flash	55.9%	estimated ± 6.0 pp, low confidence
106	MiniCPM5-2B	55.9%	estimated ± 6.0 pp, low confidence
107	Mistral Large 3	55.9%	estimated ± 6.0 pp, low confidence
108	Mistral Medium 3.5 128B	55.9%	estimated ± 6.0 pp, low confidence
109	Mistral Small 4	55.9%	estimated ± 6.0 pp, low confidence
110	Mistral Small 4 (Reasoning)	55.9%	estimated ± 6.0 pp, low confidence
111	Nemotron 3.5 Lightning 30B A3B NVFP4	55.9%	estimated ± 6.0 pp, low confidence
112	Nemotron 3 Nano 30B	55.9%	estimated ± 6.0 pp, low confidence
113	Nemotron 3 Nano Omni 30B A3B	55.9%	estimated ± 6.0 pp, low confidence
114	Nemotron 3 Super 100B	55.9%	estimated ± 6.0 pp, low confidence
115	North Mini Code	55.9%	estimated ± 6.0 pp, low confidence
116	Qwen3.5-122B-A10B	55.9%	estimated ± 6.0 pp, low confidence
117	Solar Pro 3	55.9%	estimated ± 6.0 pp, low confidence
118	Trinity-Large-Preview	55.9%	estimated ± 6.0 pp, low confidence
119	Trinity-Large-Thinking	55.9%	estimated ± 6.0 pp, low confidence
120	Ultravox v0.6 Llama 3.3 70B	55.9%	estimated ± 6.0 pp, low confidence
121	Solar Open 2	55.4%	estimated ± 2.6 pp, high confidence
122	GPT-5.4 mini	55.2%	estimated ± 2.6 pp, high confidence
123	GPT-5.4 nano	54.9%	estimated ± 2.6 pp, high confidence
124	Kimi K2.6	54.9%	estimated ± 2.6 pp, high confidence
125	Ornith-1.5-9B	54.6%	estimated ± 2.6 pp, high confidence
126	Kimi K2.5	54.3%	measured
127	GLM-5	54.1%	measured
128	LLaDA2.2-flash	53.7%	estimated ± 2.6 pp, high confidence
129	Qwen3.6-27B	53.4%	measured
130	Qwen3.6-35B-A3B	52.6%	measured
131	Claude Opus 4.5	52.3%	measured
132	Agents-A1-4B	52.0%	estimated ± 2.0 pp, medium confidence
133	Agents-A1	52.0%	estimated ± 2.0 pp, medium confidence
134	Claude Sonnet 4.5	52.0%	estimated ± 2.0 pp, low confidence
135	DeepSeek V3.2	52.0%	estimated ± 2.0 pp, low confidence
136	GLM-4.7	52.0%	estimated ± 2.0 pp, low confidence
137	LongCat-Flash-Lite-Sparse	52.0%	estimated ± 2.0 pp, low confidence
138	Qwen3.5 397B	51.8%	measured

QwenClawBench leaderboard

All leaderboards

Agentic · terminal

Agentic · tools

Coding

Math

Knowledge & reasoning

Instruction following

Multilingual

Vision & documents

Long context

General