Benchmarks · coding
Terminal-Bench 2.0 (Vals)
State-of-the-art set of difficult terminal-based tasks Scores are the values published by VALS AI. Source benchmark version 2.0.
27 models · 27 results · updated 2026-06-04 · vals.ai ↗
Leaderboard
GPT-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. 14 more models under All models in settings.
More from this benchmark
4 published measuresTime per task65 results · lower is better
16 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 | |||
|---|---|---|---|
| GPT-5.5 high | 85.0 | 79.3 | 60.5 |
| Opus 4.8 max | 91.7 | 77.0 | 54.4 |
| Opus 4.7 high | 100.0 | 72.7 | 56.7 |
| Gemini 3.5 Flash high | 100.0 | 72.7 | 53.3 |
| GPT-5.3 Codex high | 100.0 | 69.1 | 50.0 |
| Muse Spark | 75.0 | 69.1 | 40.0 |
| Sonnet 4.6 max | 75.0 | 69.1 | 40.0 |
| Qwen 3.7 Max | 83.3 | 67.9 | 40.0 |
| Kimi K2.6 max | 75.0 | 67.3 | 36.7 |
| GLM 5.1 max | 75.0 | 61.8 | 36.7 |
| MiniMax M2.7 high | 75.0 | 56.4 | 26.7 |
| MiniMax M3 max | 50.0 | 60.0 | 20.0 |
| Qwen 3.6 27B max | 75.0 | 56.4 | 20.0 |
| Qwen 3.6 Plus max | 75.0 | 52.7 | 26.7 |
| Grok 4.3 xhigh | 66.7 | 52.1 | 24.4 |
| MiniMax M2.5 high | 75.0 | 52.7 | 16.7 |
| Gemma 4 31B IT high | 75.0 | 45.5 | 23.3 |
| GLM 4.7 none | 75.0 | 43.6 | 23.3 |
| MiniMax M2.1 | 50.0 | 49.1 | 13.3 |
| Laguna M.1 | 75.0 | 38.2 | 13.3 |
| Gemini 2.5 Pro high | 100.0 | 38.2 | 6.67 |
| Mistral Medium 3.5 high | 75.0 | 41.8 | 3.33 |
| GLM 4.6 none | 75.0 | 34.5 | 10.0 |
| Laguna XS.2 | 75.0 | 34.5 | 10.0 |
| Qwen 3 Max | 75.0 | 29.1 | 10.0 |
All results
27 rows| # | Model | Effort | Score | Subscription $ / task | API $ / task | Time |
|---|---|---|---|---|---|---|
| 1 | GPT-5.5 | high | 73.2% | $0.021 | $0.76 | 7.7 min |
| 2 | Opus 4.8 | max | 70.0% | $0.015 | $0.88 | 7.9 min |
| 3 | Opus 4.7 | high | 68.5% | $0.98 | $0.98 | 11 min |
| 4 | Gemini 3.5 Flash | high | 67.4% | $1.19 | $1.19 | 6.8 min |
| 5 | GPT-5.3 Codex | high | 64.0% | $0.35 | $0.35 | 8.4 min |
| 6 | Muse Spark | 59.6% | $0.070 | $0.070 | 9.4 min | |
| 7 | Sonnet 4.6 | max | 59.6% | $0.52 | $0.52 | 10 min |
| 8 | Qwen 3.7 Max | max | 59.2% | $0.22 | $0.72 | 11 min |
| 9 | Kimi K2.6 | max | 57.3% | $0.096 | $0.19 | 13 min |
| 10 | GLM 5.1 | max | 53.9% | $0.041 | $0.16 | 15 min |
| 11 | MiniMax M2.7 | high | 47.2% | $0.013 | $0.074 | 14 min |
| 12 | MiniMax M3 | max | 46.1% | $0.036 | $0.22 | 17 min |
| 13 | Qwen 3.6 27B | max | 44.9% | $0.30 | $0.30 | 33 min |
| 14 | Qwen 3.6 Plus | max | 44.9% | $0.072 | $0.15 | 14 min |
| 15 | Grok 4.3 | xhigh | 43.4% | $0.37 | $0.37 | 11 min |
| 16 | MiniMax M2.5 | high | 41.6% | $0.043 | $0.085 | 13 min |
| 17 | Gemma 4 31B IT | high | 39.3% | 14 min | ||
| 18 | GLM 4.7 | none | 38.2% | $0.18 | $0.18 | 13 min |
| 19 | MiniMax M2.1 | 37.1% | $0.070 | $0.070 | 13 min | |
| 20 | Laguna M.1 | 31.5% | ||||
| 21 | Gemini 2.5 Pro | high | 30.3% | $0.47 | $0.47 | 13 min |
| 22 | Mistral Medium 3.5 | high | 30.3% | $0.22 | $4.71 | 21 min |
| 23 | GLM 4.6 | none | 28.1% | $0.14 | $0.14 | 9.6 min |
| 24 | Laguna XS.2 | 28.1% | ||||
| 25 | Qwen 3 Max | max | 24.7% | $0.42 | $0.42 | 15 min |
| 26 | Qwen 3.5 Flash | max | 24.7% | $0.11 | $0.11 | 12 min |
| 27 | gpt-oss-120b | high | 19.1% | $0.13 | $0.13 | 8.2 min |