MapSatisfyBench - Factual Faithfulness: leaderboard
MapSatisfyBench (2026) gives map-service agents informal, underspecified requests rebuilt from real user behavior and checks whether they recover the unstated needs through 22 replayed map tools and a simulated user, with GPT-5.3 as judge. This board is the percentage of required factual claims in the answer that the tool outputs support; higher is better.
Metric: Factual-faithfulness rate (%): share of required factual claims in the final response supported by the tool outputs; underspecified map-service requests rebuilt from real anonymized user behavior chains, answered by a ReAct agent with 22 deterministic replayed map tools and a simulated user; GPT-5.3 simulates the user and judges (majority of three judgments), temperature 1; higher is better. Source: arxiv.org. Saturation forecast: Estimated already saturated. 12 models tracked. Updated 30 Sep 2026.
Top models
| # | Model | Setting | Score | Overall rank |
|---|---|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | Non-reasoning | 96.45 | #93 |
| 2 | Gemini 3.1 Flash (Preview) | Non-reasoning | 94.48 | #559 |
| 3 | GPT-4.1 | 92.15 | #500 | |
| 4 | Qwen 3 30B A3B | Non-reasoning | 92.14 | #1147 |
| 5 | Claude Sonnet 4.6 | 91.93 | #165 | |
| 6 | Qwen 3 8B | Non-reasoning | 91.83 | #1868 |
| 7 | Qwen 3.6 Plus | Thinking | 90.84 | #226 |
| 8 | Claude Opus 4.6 | 90.77 | #106 | |
| 9 | DeepSeek V4 Pro | Thinking | 90.69 | #191 |
| 10 | GPT-5.3 | Non-reasoning | 89.62 | #197 |
| 11 | DeepSeek V3.2 | Non-reasoning | 88.25 | #412 |
| 12 | Qwen 3 235B A22B | Non-reasoning | 83.77 | #628 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=mapsatisfybench-factual-faithfulness · How It Works · Data refreshed daily, snapshot 2026-11-16.