Benchmarks · agentic
Terminal-Bench 4.0 (Vals)
Vals AI's Terminal-Bench 4.0 aggregate scores from its external-runner leaderboard. Vals publishes a separate system set and score series from the tbench.ai harness-plus-model leaderboard, so the two sources remain distinct.
43 models · 44 results · updated 2026-10-27 · vals.ai ↗
Leaderboard
Opus 5.5 max
More from this benchmark
1 published measuresTime per task57 results · lower is better
All results
44 rows| # | Model | Effort | Score | Subscription $ / task | API $ / task | Time |
|---|---|---|---|---|---|---|
| 1 | Opus 5.5 | max | 65.2% | $0.20 | $13 | 1.1 h |
| 2 | Sonnet 5.5 | max | 64.1% | $0.25 | $17 | 1.4 h |
| 3 | Gemini 4 Carbon | high | 63.1% | $7.45 | $7.45 | 45 min |
| 4 | GPT-6 Astra | max | 59.6% | $0.82 | $9.58 | 36 min |
| 5 | Fable 5.1 | max | 58.1% | $2.68 | $17 | 1.0 h |
| 6 | Gemini 4 Argon | high | 57.6% | $6.29 | $18 | 54 min |
| 7 | GPT-6.1 Sol | max | 55.1% | $0.20 | $1.72 | 45 min |
| 8 | Opus 5 | max | 53.5% | $0.32 | $19 | 1.1 h |
| 9 | GPT-6.1 Terra | max | 49.5% | $2.04 | $2.04 | 33 min |
| 10 | GPT-6 Sol | max | 44.4% | $0.36 | $5.79 | 35 min |
| 11 | Fable 5 | max | 41.4% | $3.42 | $30 | 1.1 h |
| 12 | GLM 5.3 | max | 38.9% | $0.97 | $9.37 | 1.4 h |
| 13 | GPT-5.6 Sol | max | 37.9% | $0.44 | $7.98 | 34 min |
| 14 | GLM 5.3 Flash | max | 37.4% | free | $1.22 | 1.0 h |
| 15 | Haiku 5.5 | max | 36.4% | $0.066 | $3.12 | 52 min |
| 16 | Qwen 3.8 Max | max | 34.3% | $11 | $11 | 2.5 h |
| 17 | DeepSeek V4.1 Flash 1017 | max | 31.8% | free | $0.54 | 41 min |
| 18 | MiMo V2.6 Pro | high | 31.3% | $0.38 | $0.46 | 2.7 h |
| 19 | Grok 4.7 | xhigh | 28.8% | $1.67 | $18 | 50 min |
| 20 | Mistral Large 4 | high | 27.8% | $10 | $10 | 1.7 h |
| 21 | Muse Spark 1.3 | max | 24.7% | $3.33 | $6.65 | 53 min |
| 22 | MiMo V2.6 Flash | high | 24.2% | free | $0.21 | 2.4 h |
| 23 | Opus 4.8 | max | 23.2% | $0.30 | $17 | 1.2 h |
| 24 | GPT-5.6 Terra | max | 22.7% | $0.31 | $5.60 | 31 min |
| 25 | DeepSeek V4.1 Flash | high | 19.7% | $0.031 | $0.50 | 48 min |
| 26 | Gemini 3.8 Flash | high | 19.7% | $0.35 | $8.65 | 1.8 h |
| 27 | DeepSeek V4 Flash 0731 | high | 18.7% | $0.77 | $0.77 | 1.8 h |
| 28 | Grok 4.6 | high | 17.2% | $0.52 | $5.06 | 29 min |
| 29 | Kimi K3 | max | 17.2% | $0.96 | $12 | 3.2 h |
| 30 | DeepSeek V4 Pro 0813 | max | 14.1% | $3.31 | $3.31 | 1.5 h |