Agentic · tools

LifeVal leaderboard

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

As of 2026-11-03, the highest measured score on LifeVal is 80.6% by GPT-6 Astra. 53 more models have estimated scores, calibrated from the benchmarks they were measured on.

Measured scores: benchlm.ai.

#	Model	Score	Source
1	Atria Dawn Preview	80.9%	estimated ± 1.6 pp, low confidence
2	Kimi K3.1 Pro	80.7%	estimated ± 1.6 pp, medium confidence
3	GPT-6 Astra	80.6%	measured
4	Ling 3.1 Flash	80.5%	estimated ± 1.6 pp, medium confidence
5	Kimi K3	80.2%	estimated ± 1.6 pp, medium confidence
6	Claude Opus 5	80.0%	estimated ± 1.6 pp, medium confidence
7	GPT-5.6 Sol	80.0%	measured
8	GPT-5.5 Pro	79.6%	estimated ± 1.6 pp, medium confidence
9	GPT-5.4 Pro	79.1%	estimated ± 1.6 pp, medium confidence
10	Step 5 Preview	78.8%	estimated ± 1.6 pp, medium confidence
11	Claude Mythos 5	78.3%	estimated ± 1.6 pp, medium confidence
12	GPT-5.6 Terra	78.0%	estimated ± 1.6 pp, medium confidence
13	GPT-5.5	77.4%	measured
14	Ornith-1.5-397B	77.3%	estimated ± 1.6 pp, medium confidence
15	Claude Sonnet 5	75.9%	estimated ± 1.6 pp, medium confidence
16	Mistral Large 4	75.7%	estimated ± 1.6 pp, medium confidence
17	Claude Opus 4.8	75.6%	estimated ± 1.6 pp, medium confidence
18	Claude Opus 4.6	75.0%	estimated ± 1.6 pp, medium confidence
19	MiniMax M3	74.9%	estimated ± 1.6 pp, medium confidence
20	DeepSeek V4 Pro 0813	74.8%	estimated ± 1.6 pp, medium confidence
21	dots3-note Preview	74.7%	estimated ± 1.6 pp, medium confidence
22	GPT-5.6 Luna	74.7%	estimated ± 1.6 pp, medium confidence
23	Kimi K2.6	74.6%	estimated ± 1.6 pp, medium confidence
24	Sage-A1.5	73.2%	estimated ± 1.6 pp, medium confidence
25	GPT-5.4	73.0%	measured
26	Ling 3.0 Flash Sante	71.7%	estimated ± 1.6 pp, medium confidence
27	Claude Opus 4.7 (Adaptive)	70.7%	estimated ± 1.6 pp, medium confidence
28	Nemotron 3.5 Super VL 120B A12B	69.2%	estimated ± 1.6 pp, medium confidence
29	Nemotron 3.5 Super VL 120B A12B NVFP4	68.6%	estimated ± 1.6 pp, medium confidence
30	Inkling-Small	68.5%	estimated ± 1.6 pp, medium confidence
31	Inkling	68.1%	estimated ± 1.6 pp, medium confidence
32	Beam 1.0 501B-A23B	67.7%	estimated ± 1.6 pp, medium confidence
33	Beam 1.0 501B-A23B FP8	67.6%	estimated ± 1.6 pp, medium confidence
34	Step 3.7 Flash	66.4%	estimated ± 1.6 pp, medium confidence
35	Beam 1.0 501B-A23B NVFP4	66.1%	estimated ± 1.6 pp, medium confidence
36	Agents-A1	66.0%	estimated ± 1.6 pp, medium confidence
37	DeepSeek V4 Flash 0731	62.6%	estimated ± 1.6 pp, medium confidence
38	Ling 3.0 Flash	61.1%	estimated ± 1.6 pp, medium confidence
39	Nemotron 3.5 Bolt 74B-A9B BF16	57.3%	estimated ± 1.6 pp, medium confidence
40	Qwen3.5 397B	55.5%	estimated ± 1.6 pp, medium confidence
41	GLM-5.1	53.7%	estimated ± 1.6 pp, medium confidence
42	Ornith-1.5-35B-A3B	52.9%	estimated ± 1.6 pp, medium confidence
43	Agents-A1-4B	51.4%	estimated ± 1.6 pp, medium confidence
44	Nemotron 3.5 Bolt 74B-A9B	51.3%	estimated ± 1.6 pp, medium confidence
45	GPT-5.2	49.3%	measured
46	Qwen3.5-122B-A10B	45.3%	estimated ± 1.6 pp, low confidence
47	GLM-5	41.4%	estimated ± 1.6 pp, low confidence
48	Qwen3.5-27B	39.3%	estimated ± 1.6 pp, low confidence
49	Qwen3.5-35B-A3B	39.3%	estimated ± 1.6 pp, low confidence
50	Kimi K2.5	38.4%	estimated ± 1.6 pp, low confidence
51	Kimi K2.5 (Reasoning)	38.4%	estimated ± 1.6 pp, low confidence
52	Ornith-1.5-9B	29.4%	estimated ± 1.6 pp, low confidence
53	Spark-X2.6-4B	22.7%	estimated ± 1.6 pp, low confidence
54	GLM-4.7	20.7%	estimated ± 1.6 pp, low confidence
55	Sarvam 105B	16.3%	estimated ± 1.6 pp, low confidence
56	Solar Pro 4	15.9%	estimated ± 1.6 pp, low confidence
57	LongCat-Flash-Lite-Sparse	15.0%	estimated ± 1.6 pp, low confidence
58	Nemotron 3 Ultra	9.3%	estimated ± 1.6 pp, low confidence

LifeVal leaderboard

All leaderboards

Agentic · terminal

Agentic · tools

Coding

Math

Knowledge & reasoning

Instruction following

Multilingual

Vision & documents

Long context

General