Self-Driving Negotiator: leaderboard
Metric: Average success rate (%). Source: arxiv.org. 6 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 2.5 Flash | 67 |
| 2 | Llama 3.3 70B Instruct | 62 |
| 3 | Mistral Small 3.2 | 56 |
| 4 | DeepSeek V3.2 | 56 |
| 5 | Llama 3.2 3B Instruct | 47 |
Interactive version: theaggregate.ai/benchmark?slug=self-driving-negotiator · How It Works · Data refreshed daily, snapshot 2026-09-05.