DORA - Temporal Evolution Reasoning: leaderboard
Metric: Final-answer accuracy (%; ReAct agent with tools). Source: github.com. Saturation forecast: Around December 2026. 13 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | MiMo-V2-Pro | 63.58 |
| 2 | Qwen 3.5 397B A17B | 61.83 |
| 3 | Gemma 4 31B | 61.03 |
| 4 | Gemini 3 Flash | 60.4 |
| 5 | Claude Sonnet 4.6 | 60.05 |
| 6 | Grok 4.1 Fast | 55.28 |
| 7 | GPT-5.4 | 51.9 |
| 8 | DeepSeek V3.2 | 50.62 |
| 9 | MiniMax-M2.7 | 50.59 |
| 10 | Step 3.5 Flash | 48.9 |
| 11 | GPT-5.4 Nano | 45.55 |
| 12 | Qwen 3.5 35B A3B | 37.62 |
| 13 | GPT-OSS-120B | 24.43 |
Interactive version: theaggregate.ai/benchmark?slug=dora-temporal-evolution-reasoning · How It Works · Data refreshed daily, snapshot 2026-10-02.