Galileo Agent - Insurance TSQ: leaderboard
Task Success Quality score for Galileo insurance workflows, measuring the quality of multi-step policy and claims assistance.
Metric: Task Success Quality (%). Source: huggingface.co. Saturation forecast: Estimated already saturated. 22 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 2.5 Flash | 94 | #245 |
| 2 | GLM 4.5 Air | 94 | #392 |
| 3 | Claude Sonnet 4 (20250514) | 93 | #213 |
| 4 | Grok 4 | 91 | #181 |
| 5 | Kimi K2 | 88 | #243 |
| 6 | Gemini 2.5 Pro | 87 | #152 |
| 7 | Qwen 3 235B A22B 2507 (Thinking) | 86 | #260 (Qwen 3 235B A22B 2507) |
| 8 | Qwen 3 235B A22B 2507 Instruct | 85 | #300 |
| 9 | Gemini 2.5 Flash Lite | 78 | #422 |
| 10 | Qwen 2.5 72B Instruct | 77 | #454 |
| 11 | DeepSeek V3 | 76 | #321 |
| 12 | Mistral Medium 3.1 | 73 | #408 |
| 13 | GPT-4.1 | 68 | #248 |
| 14 | Mistral Small 3.2 | 65 | #605 |
| 15 | Nova Pro (v1) | 64 | #497 |
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-tsq · How It Works · Data refreshed daily, snapshot 2026-10-20.