FinSearchComp — combined Tier 2 + Tier 3 — Combined-tier aggregate, distinct subset scope from standalone T3.
| # | Model | Vendor | Best score | Runs | Last seen |
|---|---|---|---|---|---|
| 1 | Kimi K2.5 | Moonshot | 67.8 | 2 | 2026-08-23 |
| 2 | Claude Opus 4.5 | Anthropic | 66.2 | 1 | 2026-06-15 |
| 3 | DeepSeek-V3.2 | DeepSeek | 59.1 | 1 | 2026-06-15 |
| 4 | Gemini 3 Pro | 49.9 | 1 | 2026-06-15 |
Best tracked score per model (default configuration; source-attributed and verification-tiered). Open a model for its full benchmark surface, provenance and pricing.