TEBench - Stale Test Identification: leaderboard

Identification F1 (%) for tests that still pass but no longer check the changed behaviour on TEBench (2026), a project-level test-evolution benchmark of 314 commits from 10 Java projects in which an agent in its default configuration updates the test suite after a production-code change. Higher is better.

Metric: Identification F1 (%) for tests left stale by the change: the tests the agent modified, deleted (method level) or added (file level) against the developer's own test changes, on TEBench (314 commits from 10 Java projects): given the project at the commit with its production changes but not its test changes, the agent edits the test suite in an isolated worktree with default agent settings; higher is better. Source: arxiv.org. Saturation forecast: Around October 2027. 6 models tracked. Updated 7 Oct 2026.

Top models

#ModelScoreOverall rank
1GPT-5.3 Codex37.4#125
2GLM-537.1#276
3Qwen3.5 (TEBench checkpoint unspecified)36
4Kimi K2.535.8#272
5Claude Sonnet 4.635.6#179
6DeepSeek V3.233.4#425

No result here: #2 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.

Interactive version: theaggregate.ai/benchmark?slug=tebench-stale-test-identification · How It Works · Data refreshed daily, snapshot 2026-11-22.