Agent2 RL-Bench - GSM8K: leaderboard
Metric: GSM8K test accuracy (%) of the Qwen3-8B-Base model after the coding agent (driver LLM in a CLI scaffold) post-trains it for 12 hours in an isolated GPU workspace, best score the grading server returned over the run, single run; higher is better. Source: arxiv.org. Saturation forecast: Estimated already saturated. 6 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5.2 | 89.46 |
| 2 | Claude Opus 4.6 | 88.78 |
| 3 | Claude Sonnet 4.5 | 88.7 |
| 4 | GPT-5.4 | 84 |
| 5 | GPT-4o | 83.32 |
| 6 | Gemini 2.5 Flash | 70.13 |
Interactive version: theaggregate.ai/benchmark?slug=agent2-rl-bench-gsm8k · How It Works · Data refreshed daily, snapshot 2026-10-07.