FinSearchComp — Global split — Financial search-and-reasoning benchmark, global-market subset.
| # | Model | Vendor | Best score | Runs | Last seen |
|---|---|---|---|---|---|
| 1 | MiniMax M2 | MiniMax | 65.5 | 1 | 2026-06-06 |
| 2 | GPT-5 | OpenAI | 63.9 | 1 | 2026-06-06 |
| 3 | Claude Sonnet 4.5 | Anthropic | 60.8 | 1 | 2026-06-06 |
| 4 | Gemini 2.5 Pro | 42.6 | 1 | 2026-06-06 | |
| 5 | Claude Sonnet 4 | Anthropic | 42 | 1 | 2026-06-06 |
| 6 | Kimi K2 (Reasoning) | Moonshot | 29.5 | 1 | 2026-06-06 |
| 7 | GLM-4.6 | Z.ai | 29.2 | 1 | 2026-06-06 |
| 8 | DeepSeek-V3.2 | DeepSeek | 26.2 | 1 | 2026-06-06 |
Best tracked score per model (default configuration; source-attributed and verification-tiered). Open a model for its full benchmark surface, provenance and pricing.