Skip to content

Benchmark

Vals Terminal-Bench 4.0

Agentic benchmarks measure how well models carry out tasks that take several steps. Scores are percentages.

Models measured
65
Scores recorded
65
Sources
1
IQ weight
2.05

Score distribution

28.6%
Mean
24.2%
Median
21.2%
Std dev
2.5%
10th percentile
60.6%
90th percentile
0%
Min
0%65.2%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 2.05 for chat models, measured over 65 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
1Claude Fable 5.2· Vals AI65.2%
2Claude Opus 5.5· Vals AI65.2%
3Claude Sonnet 5.5· Vals AI64.1%
4Kimi K3.1 Pro· Vals AI63.6%
5Gemini 4 Carbon· Vals AI63.1%
6GPT-6.2 Astra· Vals AI63.1%
7GPT-6.2 Sol· Vals AI60.6%
8GPT-6 Astra· Vals AI59.6%
9Claude Fable 5.1· Vals AI58.1%
10Muse Wonder Full· Vals AI58.1%
11Gemini 4 Argon· Vals AI57.6%
12GPT-6.1 Sol· Vals AI55.1%
13Gemini 4 Hydrogen· Vals AI54%
14Claude Opus 5· Vals AI53.5%
15GPT-6.1 Terra· Vals AI49.5%
16Muse Wonder· Vals AI49.5%
17Qwen4 Max· Vals AI48.5%
18GPT-6 Sol· Vals AI44.4%
19Claude Fable 5· Vals AI41.4%
20Hy 4· Vals AI39.4%
21Muse Ember· Vals AI39.4%
22GLM 5.3· Vals AI38.9%
23GPT-5.6 Sol· Vals AI37.9%
24GLM 5.3 Flash· Vals AI37.4%
25Claude Haiku 5.5· Vals AI36.4%
26Qwen3.8 Max· Vals AI34.3%
27DeepSeek V4.1 Flash 1017· Vals AI31.8%
28Step 5 Preview· Vals AI31.8%
29MiMo V2.6 Pro· Vals AI31.3%
30Grok 4.7· Vals AI28.8%
31Mistral Large 4· Vals AI27.8%
32MiniMax M3.1 Flash· Vals AI25.3%
33MiMo V2.6 Flash· Vals AI24.2%
34Claude Opus 4.8· Vals AI23.2%
35GPT-5.6 Terra· Vals AI22.7%
36Ember-1· Vals AI20.7%
37DeepSeek V4.1 Flash· Vals AI19.7%
38Gemini 3.8 Flash· Vals AI19.7%
39DeepSeek V4 Flash 0731· Vals AI18.7%
40Grok 4.6· Vals AI17.2%
41Kimi K3· Vals AI17.2%
42Hy4 Flash· Vals AI15.2%
43DeepSeek V4 Pro 0813· Vals AI14.1%
44GPT-6 Luna· Vals AI13.6%
45Inkling 2 Small· Vals AI12.6%
46Gemini 3.7 Flash· Vals AI12.1%
47GPT-5.6 Luna· Vals AI11.6%
48DeepSeek V4 Pro· Vals AI11.1%
49Claude Sonnet 5· Vals AI10.6%
50Muse Spark 1.3· Vals AI10.6%

Showing 1–50 of 65 results.