TensorBench - Scheduler: leaderboard
TensorBench (2026) tests coding agents (a published CLI harness with its model) on feature-addition and refactoring tasks in Scorch, a compiler-based tensor framework with dense and sparse tensors; a task passes when the patched repository passes its randomized regression tests and the tests the agent adds. This board is the pass rate in percent on the 37 scheduler tasks, cost-model-driven optimization passes such as loop transformations; higher is better.
Metric: Pass rate (%) on the 37 scheduler tasks (cost-model-driven optimization passes such as loop transformations); each coding agent (a published CLI scaffold with its model) edits the open-source Scorch compiler-based tensor framework inside a per-task container, one run per task; a task passes when every test of the post-patch suite (pre-existing randomized regression tests plus the tests the agent adds) succeeds; higher is better. Source: arxiv.org. Saturation forecast: Around May 2027. 7 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Opus 4.7 (xHigh) | 73 | #79 (Claude Opus 4.7) |
| 2 | GPT-5.5 (xHigh) | 51.4 | #49 (GPT-5.5) |
| 3 | Qwen 3 Coder 480B A35B Instruct | 29.7 | #643 |
| 4 | GPT-5.4 (xHigh) | 27 | #123 (GPT-5.4) |
| 5 | Claude Opus 4.6 | 24.3 | #102 |
| 6 | Gemini 3.1 Pro (Preview) | 18.9 | #94 |
| 7 | GPT-5.3 Codex (xHigh) | 8.1 | #116 (GPT-5.3 Codex) |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=tensorbench-scheduler · How It Works · Data refreshed daily, snapshot 2026-11-13.