Benchmarks · coding
Terminal-Bench 2.1 (Vals)
State-of-the-art set of difficult terminal-based tasks Scores are the values published by VALS AI. Source benchmark version 2.1.
58 models · 59 results · updated 2026-10-01 · vals.ai ↗
Leaderboard
Opus 5.5 high
Heavy line: best score at each price. Thin lines join one model's effort levels. Cost uses each model's median subscription plan, or API price with no plan data. 44 more models under All models in settings.
More from this benchmark
4 published measuresTime per task74 results · lower is better
48 more models under All models in settings.
Score vs any published measure3 measures: sub-scores, splits, counts
Breakdown
darker is better within each column · click a column to sort| Model | |||
|---|---|---|---|
| Opus 5.5 high | 100.0 | 90.3 | 81.1 |
| GPT-6 Astra max | 100.0 | 86.1 | 87.8 |
| GPT-5.6 Sol max | 91.7 | 86.1 | 84.4 |
| Fable 5.1 high | 100.0 | 85.5 | 82.2 |
| Opus 5 high | 100.0 | 83.6 | 84.4 |
| Sonnet 5.5 high | 91.7 | 86.1 | 76.7 |
| GPT-6 Sol max | 100.0 | 84.2 | 78.9 |
| Gemini 3.8 Flash high | 100.0 | 84.8 | 72.2 |
| Kimi K3 max | 91.7 | 83.6 | 74.4 |
| Fable 5 max | 100.0 | 84.8 | 70.0 |
| GPT-5.6 Luna max | 83.3 | 83.6 | 70.0 |
| Muse Spark 1.3 max | 100.0 | 81.2 | 72.2 |
| Grok 4.6 high | 75.0 | 83.6 | 68.9 |
| Gemini 3.7 Flash high | 91.7 | 83.6 | 64.4 |
| GPT-5.6 Terra max | 100.0 | 79.4 | 71.1 |
| GPT-5.5 high | 83.3 | 81.2 | 66.7 |
| MiMo V2.6 Flash high | 83.3 | 75.2 | 77.8 |
| Sonnet 5 max | 100.0 | 81.8 | 57.8 |
| DeepSeek V4.1 Flash high | 91.7 | 83.0 | 56.7 |
| Gemini 3.5 Flash high | 100.0 | 80.0 | 60.0 |
| Gemini 3.6 Flash high | 100.0 | 77.0 | 64.4 |
| Grok 4.7 xhigh | 91.7 | 75.2 | 67.8 |
| GPT-6 Luna max | 91.7 | 75.8 | 65.6 |
| Opus 4.8 max | 100.0 | 76.4 | 60.0 |
| GLM 5.3 max | 91.7 | 80.0 | 53.3 |
All results
59 rows| # | Model | Effort | Score | Subscription $ / task | API $ / task | Time | Harness |
|---|---|---|---|---|---|---|---|
| 1 | Opus 5.5 | high | 87.6% | $0.008 | $0.50 | 6.4 min | |
| 2 | GPT-6 Astra | max | 87.3% | $0.057 | $1.34 | 4.6 min | |
| 3 | GPT-5.6 Sol | max | 85.8% | $0.028 | $1.02 | 6.2 min | |
| 4 | Fable 5.1 | high | 85.0% | $0.48 | $3.07 | 8.3 min | |
| 5 | Opus 5 | high | 84.6% | $0.015 | $0.89 | 13 min | |
| 6 | Sonnet 5.5 | high | 83.1% | $0.009 | $0.63 | 7.8 min | |
| 7 | GPT-6 Sol | max | 83.1% | $0.012 | $0.37 | 5.0 min | |
| 8 | Gemini 3.8 Flash | high | 81.3% | $0.063 | $1.54 | 9.2 min | |
| 9 | Kimi K3 | max | 80.9% | $0.031 | $0.34 | 10 min | |
| 10 | Fable 5 | max | 80.5% | $0.16 | $1.43 | 8.4 min | |
| 11 | GPT-5.6 Luna | max | 79.0% | $0.003 | $0.054 | 5.9 min | |
| 12 | Muse Spark 1.3 | max | 79.0% | $0.30 | $0.60 | 12 min | |
| 13 | Grok 4.6 | high | 78.3% | $0.047 | $0.45 | 11 min | |
| 14 | Gemini 3.7 Flash | high | 77.5% | $0.13 | $1.02 | 8.2 min | |
| 15 | GPT-5.6 Terra | max | 77.5% | $0.013 | $0.47 | 8.7 min | |
| 16 | GPT-5.5 | high | 76.4% | $0.020 | $0.74 | 7.1 min | |
| 17 | MiMo V2.6 Flash | high | 76.4% | free | $0.023 | 13 min | |
| 18 | Sonnet 5 | max | 74.5% | $0.009 | $0.53 | 11 min | |
| 19 | DeepSeek V4.1 Flash | high | 74.5% | $0.007 | $0.099 | 10 min | |
| 20 | Gemini 3.5 Flash | high | 74.2% | $0.87 | $0.87 | 6.2 min | |
| 21 | Gemini 3.6 Flash | high | 73.8% | $0.037 | $1.17 | 9.2 min | |
| 22 | Grok 4.7 | xhigh | 73.4% | $0.100 | $1.08 | 13 min | |
| 23 | GPT-6 Luna | max | 73.0% | $0.002 | $0.029 | 7.1 min | |
| 24 | Muse Spark 1.3 | xhigh | 72.3% | $0.33 | $0.66 | 14 min | |
| 25 | Opus 4.8 | max | 71.9% | $0.042 | $2.41 | 15 min | |
| 26 | GLM 5.3 | max | 71.5% | $0.032 | $0.31 | 13 min | |
| 27 | Muse Spark 1.2 | xhigh | 69.7% | $0.25 | $0.50 | 15 min | |
| 28 | Muse Spark 1.1 | xhigh | 69.3% | $0.42 | $0.42 | 10 min | |
| 29 | Opus 4.7 | high | 68.5% | $1.98 | $1.98 | 14 min | |
| 30 | GLM 5.2 | max | 67.8% | $0.013 | $0.43 | 14 min |