ZeroBench — Kimi K2.5 card; tool-augmented run, distinct tools axis from base ZeroBench
| # | Model | Vendor | Best score | Runs | Last seen |
|---|---|---|---|---|---|
| 1 | Gemini 3 Pro | 12 | 1 | 2026-06-15 | |
| 2 | Kimi K2.5 | Moonshot | 11 | 1 | 2026-06-15 |
| 3 | Claude Opus 4.5 | Anthropic | 9 | 1 | 2026-06-15 |
| 4 | GPT-5.2 | OpenAI | 7 | 1 | 2026-06-15 |
| 5 | Qwen3 VL 235B A22B Reasoning | Alibaba | 3 | 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.