Qwen3.7 Launch - Terminal Bench 2.0-Terminus: leaderboard
The "Terminal Bench 2.0-Terminus" subtask of Qwen3.7 Launch. A benchmark result reported in the Qwen3.7 launch announcement, as self-reported by the model developer.
Metric: Score (%). Source: qwen.ai. Saturation forecast: Around January 2027. 6 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen 3.7 Max | 69.7 | #75 |
| 2 | DeepSeek V4 Pro (Reasoning, Max Effort) | 67.9 | #102 (DeepSeek V4 Pro) |
| 3 | Kimi K2.6 (Thinking) | 66.7 | #104 (Kimi K2.6) |
| 4 | Claude Opus 4.6 (Thinking) | 65.4 | #63 (Claude Opus 4.6) |
| 5 | GLM-5.1 (Thinking) | 63.5 | #97 (GLM-5.1) |
| 6 | Qwen 3.6 Plus | 61.6 | #120 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=qwen3-7-launch-terminal-bench-2-0-terminus · How It Works · Data refreshed daily, snapshot 2026-10-20.