Galileo Agent - Insurance Accuracy: leaderboard
Galileo agent benchmark slice for insurance workflows, measuring correct handling of policy, claims, coverage, and customer-support tasks.
Metric: Accuracy (%). Source: huggingface.co. Saturation forecast: Around January 2027. 22 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen 3 235B A22B 2507 Instruct | 74 | #300 |
| 2 | Mistral Medium 3.1 | 70 | #408 |
| 3 | GPT-4.1 | 66 | #248 |
| 4 | Kimi K2 | 58 | #243 |
| 5 | Gemini 2.5 Pro | 54 | #152 |
| 6 | Gemini 2.5 Flash Lite | 54 | #422 |
| 7 | Claude Sonnet 4 (20250514) | 53 | #213 |
| 8 | GLM 4.5 Air | 53 | #392 |
| 9 | Qwen 2.5 72B Instruct | 52 | #454 |
| 10 | Grok 4 | 48 | #181 |
| 11 | DeepSeek V3 | 48 | #321 |
| 12 | GPT-4.1 Mini | 46 | #355 |
| 13 | Gemini 2.5 Flash | 44 | #245 |
| 14 | Qwen 3 235B A22B 2507 (Thinking) | 42 | #260 (Qwen 3 235B A22B 2507) |
| 15 | GPT-4.1 Nano | 41 | #731 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=galileo-agent-insurance-accuracy · How It Works · Data refreshed daily, snapshot 2026-10-20.