Agentic · tools

Hermes Index leaderboard

Name: benchgap: LLM benchmark scores, measured and estimated
Creator: benchgap
License: https://creativecommons.org/licenses/by-nc/4.0/

As of 2026-10-29, the highest measured score on Hermes Index is 63.3% by Claude Opus 5.5. 136 more models have estimated scores, calibrated from the benchmarks they were measured on.

Measured scores: benchlm.ai.

#	Model	Score	Source
1	Claude Opus 5.5	63.3%	measured
2	Gemini 4 Carbon	61.0%	estimated ± 2.6 pp, low confidence
3	Muse Wonder Full	58.9%	estimated ± 3.1 pp, high confidence
4	GPT-6.1 Terra	58.9%	estimated ± 3.1 pp, high confidence
5	Claude Opus 5	58.9%	estimated ± 3.1 pp, high confidence
6	Claude Fable 5	58.8%	estimated ± 3.1 pp, high confidence
7	Claude Fable 5.1	58.7%	estimated ± 3.1 pp, high confidence
8	GPT-6.1 Sol	57.1%	measured
9	Gemini 4 Argon	56.5%	estimated ± 2.6 pp, medium confidence
10	GPT-6 Astra	56.3%	measured
11	GPT-5.6 Sol	53.8%	estimated ± 2.6 pp, medium confidence
12	Claude Sonnet 5.5	53.1%	measured
13	GPT-5.6 Terra	47.1%	estimated ± 2.6 pp, medium confidence
14	Qwen3.8 Max Preview	46.3%	estimated ± 8.7 pp, medium confidence
15	Kimi K3.1 Pro	44.9%	estimated ± 2.6 pp, medium confidence
16	Ling 3.1 Flash	44.9%	estimated ± 8.7 pp, medium confidence
17	Qwen3.8-Flash-Next	44.1%	estimated ± 8.1 pp, low confidence
18	GPT-6 Sol	44.1%	measured
19	MiMo-V2.6-Pro	42.0%	estimated ± 2.6 pp, medium confidence
20	Claude Mythos Preview	41.7%	estimated ± 2.6 pp, medium confidence
21	MiniMax M3.1 Flash	40.2%	estimated ± 7.7 pp, medium confidence
22	Grok 4.7	39.3%	measured
23	GLM-5.3	39.2%	measured
24	Kimi K3	39.2%	measured
25	DeepSeek V4.1 Flash 1017	39.0%	estimated ± 2.6 pp, medium confidence
26	Ling 3.0 Flash Sante	38.4%	estimated ± 10.0 pp, medium confidence
27	Gemini 3.8 Flash	38.1%	measured
28	Gemini 3.1 Pro	37.2%	estimated ± 3.1 pp, high confidence
29	Gemini 3.7 Flash	37.2%	estimated ± 3.1 pp, high confidence
30	Grok 4.6	37.2%	estimated ± 3.1 pp, high confidence
31	Claude Haiku 4.5	37.2%	estimated ± 3.1 pp, high confidence
32	Claude Opus 4.8	37.2%	estimated ± 3.1 pp, high confidence
33	Claude Sonnet 4.6	37.2%	estimated ± 3.1 pp, high confidence
34	Claude Sonnet 5	37.2%	estimated ± 3.1 pp, high confidence
35	DeepSeek V4 Flash 0731	37.2%	estimated ± 3.1 pp, high confidence
36	Gemini 3.5 Flash	37.2%	estimated ± 3.1 pp, high confidence
37	Gemini 3.6 Flash	37.2%	estimated ± 3.1 pp, high confidence
38	GLM-5.2	37.2%	estimated ± 3.1 pp, high confidence
39	GPT-5.3-Codex-Spark	37.2%	estimated ± 3.1 pp, high confidence
40	Grok 4.20	37.2%	estimated ± 3.1 pp, high confidence
41	Llama 4 Maverick	37.2%	estimated ± 3.1 pp, high confidence
42	MiMo-V2.5-Pro	37.2%	estimated ± 3.1 pp, high confidence
43	MiniMax M3	37.2%	estimated ± 3.1 pp, high confidence
44	Muse Spark 1.2	37.2%	estimated ± 3.1 pp, high confidence
45	Muse Spark 1.3	37.2%	estimated ± 3.1 pp, high confidence
46	Muse Spark Open	37.2%	estimated ± 3.1 pp, high confidence
47	Muse Wonder	37.2%	estimated ± 3.1 pp, high confidence
48	Nemotron 3 Ultra	37.2%	estimated ± 3.1 pp, high confidence
49	Qwen3.8-27B	37.2%	estimated ± 3.1 pp, high confidence
50	Grok 4.5	37.2%	estimated ± 8.7 pp, medium confidence
51	Mistral Large 4	37.1%	estimated ± 7.7 pp, medium confidence
52	Claude Haiku 5.5	37.0%	measured
53	GPT-5.5	36.4%	estimated ± 2.6 pp, medium confidence
54	Ornith-1.5-397B	36.4%	estimated ± 10.0 pp, medium confidence
55	Qwen3.8 Max	36.2%	measured
56	Claude Opus 4.7 (Adaptive)	36.1%	estimated ± 8.7 pp, medium confidence
57	GLM-5.3-Flash	34.9%	measured
58	GPT-5.6 Luna	34.9%	estimated ± 2.6 pp, medium confidence
59	Sage-A1.5	34.4%	estimated ± 10.0 pp, low confidence
60	GPT-6 Luna	33.9%	measured
61	Inkling	33.6%	estimated ± 7.7 pp, medium confidence
62	Beam 1.0 501B-A23B	33.4%	estimated ± 7.7 pp, medium confidence
63	Step 5 Preview	32.8%	measured
64	DeepSeek V4 Pro 0813	32.1%	estimated ± 8.1 pp, low confidence
65	Hy3 Preview	31.3%	estimated ± 8.7 pp, medium confidence
66	Apodex 1.1	30.2%	estimated ± 8.7 pp, medium confidence
67	Apodex 1.1 Mini	30.2%	estimated ± 8.7 pp, medium confidence
68	Quasar 438B	30.0%	estimated ± 8.7 pp, medium confidence
69	dots3-note Preview	29.4%	estimated ± 10.0 pp, low confidence
70	Nemotron 3.5 Super VL 120B A12B	29.2%	estimated ± 8.7 pp, medium confidence
71	Nemotron 3.5 Super VL 120B A12B NVFP4	29.2%	estimated ± 8.7 pp, medium confidence
72	Ling 3.0 Flash VL	29.2%	estimated ± 8.7 pp, medium confidence
73	Qwen3.7 Max	27.9%	estimated ± 8.7 pp, medium confidence
74	Inkling-Small	27.8%	estimated ± 8.7 pp, medium confidence
75	GLM-5.1	27.5%	estimated ± 8.7 pp, medium confidence
76	Solar Pro 4	27.1%	estimated ± 8.7 pp, medium confidence
77	Beam 1.0 501B-A23B FP8	26.4%	estimated ± 10.0 pp, low confidence
78	Beam 1.0 501B-A23B NVFP4	26.4%	estimated ± 10.0 pp, low confidence
79	Grok 4.3	26.2%	estimated ± 8.7 pp, medium confidence
80	Hy4 preview	26.1%	measured
81	Hy3	25.5%	estimated ± 8.7 pp, medium confidence
82	GLM-5	25.0%	estimated ± 10.0 pp, low confidence
83	Kimi K2.6	24.4%	estimated ± 8.7 pp, medium confidence
84	Kimi K2.7 Code	24.4%	estimated ± 8.7 pp, medium confidence
85	Step 3.7 Flash	23.6%	estimated ± 8.7 pp, medium confidence
86	GLM-4.7	23.5%	estimated ± 8.7 pp, medium confidence
87	GPT-5.4 mini	23.4%	estimated ± 8.7 pp, medium confidence
88	MiniMax M2.7	23.3%	estimated ± 8.7 pp, medium confidence
89	Muse Spark	22.8%	estimated ± 8.7 pp, medium confidence
90	Qwen3.6 Plus	22.5%	estimated ± 8.7 pp, medium confidence
91	Qwen3.6-27B	22.3%	estimated ± 8.7 pp, medium confidence
92	Gemini 3.5 Flash-Lite	22.3%	estimated ± 8.7 pp, medium confidence
93	Nemotron 3.5 Bolt 74B-A9B	22.1%	estimated ± 8.7 pp, medium confidence
94	GPT-5.4	21.9%	estimated ± 2.6 pp, low confidence
95	MiMo-V2.6-Flash	21.9%	estimated ± 2.6 pp, low confidence
96	Ling 3.0 Flash	21.6%	measured
97	A.X K2	21.3%	estimated ± 8.7 pp, medium confidence
98	GPT-5.4 nano	20.8%	estimated ± 8.7 pp, medium confidence
99	GPT-5 (high)	19.5%	estimated ± 8.7 pp, low confidence
100	Agents-A1	19.3%	estimated ± 10.0 pp, low confidence
101	Qwen3.6-35B-A3B	18.4%	estimated ± 8.7 pp, low confidence
102	Kimi K2.5	16.2%	estimated ± 8.7 pp, low confidence
103	Kimi K2.5 (Reasoning)	16.2%	estimated ± 8.7 pp, low confidence
104	GPT-5.1	15.6%	estimated ± 8.7 pp, low confidence
105	Qwen3.5-122B-A10B	15.0%	estimated ± 8.7 pp, low confidence
106	K-EXAONE 2.0	14.8%	estimated ± 8.7 pp, low confidence
107	Muse Glimmer 30B	13.7%	estimated ± 8.7 pp, low confidence
108	Qwen3.7 Plus	12.9%	estimated ± 8.7 pp, low confidence
109	Mistral Medium 3.5 128B	12.6%	estimated ± 8.7 pp, low confidence
110	MiniCPM5-2B	10.2%	estimated ± 8.7 pp, low confidence
111	Nemotron 3.5 Lightning 30B A3B NVFP4	6.9%	estimated ± 8.7 pp, low confidence
112	MiMo-V2-Flash	6.4%	estimated ± 8.7 pp, low confidence
113	Gemma 4 31B	5.9%	estimated ± 8.7 pp, low confidence
114	GPT-OSS 120B	5.4%	estimated ± 8.7 pp, low confidence
115	Muse Spark 1.1	3.8%	estimated ± 2.6 pp, low confidence
116	Granite 4.2 30B	3.8%	estimated ± 8.7 pp, low confidence
117	Gemma 4 26B A4B	3.3%	estimated ± 8.7 pp, low confidence
118	Ling 3.0 Tiny	3.1%	estimated ± 8.7 pp, low confidence
119	Command A+	0.7%	estimated ± 8.7 pp, low confidence
120	Celeris-1	0.0%	estimated ± 8.7 pp, low confidence
121	Claude 3.5 Haiku	0.0%	estimated ± 8.7 pp, low confidence
122	DeepSeek V3	0.0%	estimated ± 8.7 pp, low confidence
123	DeepSeek V3 0324	0.0%	estimated ± 8.7 pp, low confidence
124	Gemini 2.5 Pro	0.0%	estimated ± 8.7 pp, low confidence
125	Gemma 3 27B	0.0%	estimated ± 8.7 pp, low confidence
126	Gemma 4 12B	0.0%	estimated ± 8.7 pp, low confidence
127	Gemma 4 E2B	0.0%	estimated ± 8.7 pp, low confidence
128	Gemma 4 E4B	0.0%	estimated ± 8.7 pp, low confidence
129	GPT-4.1 mini	0.0%	estimated ± 8.7 pp, low confidence
130	GPT-4.1 nano	0.0%	estimated ± 8.7 pp, low confidence
131	GPT-4o	0.0%	estimated ± 8.7 pp, low confidence
132	GPT-4o mini	0.0%	estimated ± 8.7 pp, low confidence
133	GPT-OSS 20B	0.0%	estimated ± 8.7 pp, low confidence
134	Granite 4.2 3B	0.0%	estimated ± 8.7 pp, low confidence
135	Granite 4.2 8B	0.0%	estimated ± 8.7 pp, low confidence
136	K-Exaone	0.0%	estimated ± 8.7 pp, low confidence
137	LFM2.5-2.6B	0.0%	estimated ± 8.7 pp, low confidence
138	Ling 2.6 Flash	0.0%	estimated ± 8.7 pp, low confidence
139	Llama 3.3 70B	0.0%	estimated ± 8.7 pp, low confidence
140	Llama 4 Scout	0.0%	estimated ± 8.7 pp, low confidence
141	Mercury 2.5	0.0%	estimated ± 8.7 pp, low confidence
142	Mistral Large 3	0.0%	estimated ± 8.7 pp, low confidence
143	Mistral Small 3.1	0.0%	estimated ± 8.7 pp, low confidence
144	Mistral Small 4	0.0%	estimated ± 8.7 pp, low confidence
145	Mistral Small 4 (Reasoning)	0.0%	estimated ± 8.7 pp, low confidence
146	Nemotron 3 Nano 30B	0.0%	estimated ± 8.7 pp, low confidence
147	Nemotron 3 Nano Omni 30B A3B	0.0%	estimated ± 8.7 pp, low confidence
148	North Mini Code	0.0%	estimated ± 8.7 pp, low confidence
149	Solar Pro 3	0.0%	estimated ± 8.7 pp, low confidence
150	Trinity-Large-Preview	0.0%	estimated ± 8.7 pp, low confidence
151	Trinity-Large-Thinking	0.0%	estimated ± 8.7 pp, low confidence
152	Ultravox v0.6 Llama 3.3 70B	0.0%	estimated ± 8.7 pp, low confidence

Hermes Index leaderboard

All leaderboards

Agentic · terminal

Agentic · tools

Coding

Math

Knowledge & reasoning

Instruction following

Multilingual

Vision & documents

Long context

General