Benchmark
OSWorld-Verified
OSWorld-Verified is a verified subset of OSWorld, a scalable real computer environment for multimodal agents supporting task setup, execution-based…
- Modality
- multimodal
- Categories
- multimodal, general, agents, vision
- Openness
- unknown
- Source
- llm_stats
- Reported scores
- 24
Reported scores
llm_stats
| Model | Organization | Reported value | Reported | Evidence |
|---|---|---|---|---|
| Claude Fable 5 | Anthropic | 0.85 | 2026-06-09 | evidence |
| Claude Mythos Preview | Anthropic | 0.796 | 2026-04-07 | evidence |
| Claude Opus 4.7 | Anthropic | 0.78 | 2026-04-16 | evidence |
| Claude Opus 4.8 | Anthropic | 0.834 | 2026-05-28 | evidence |
| Claude Sonnet 5 | Anthropic | 0.812 | 2026-06-30 | evidence |
| Gemini 3.5 Flash | 0.784 | 2026-05-19 | evidence | |
| Gemini 3.5 Flash-Lite | 0.74 | 2026-07-21 | evidence | |
| Gemini 3.6 Flash | 0.83 | 2026-07-21 | evidence | |
| GPT-5.3 Codex | OpenAI | 0.647 | 2026-02-05 | evidence |
| GPT-5.4 | OpenAI | 0.75 | 2026-03-05 | evidence |
| GPT-5.4 mini | OpenAI | 0.721 | 2026-03-17 | evidence |
| GPT-5.4 nano | OpenAI | 0.39 | 2026-03-17 | evidence |
| GPT-5.5 | OpenAI | 0.787 | 2026-04-23 | evidence |
| Kimi K2.6 | Moonshot AI | 0.731 | 2026-04-20 | evidence |
| MiniMax M3 | MiniMax | 0.7006 | 2026-06-01 | evidence |
| Muse Glimmer-30B | Meta | 0.659 | 2026-08-10 | evidence |
| Muse Spark 1.1 | Meta | 0.808 | 2026-07-09 | evidence |
| Qwen3.5-122B-A10B | Qwen | 0.58 | 2026-02-24 | evidence |
| Qwen3.5-27B | Qwen | 0.562 | 2026-02-24 | evidence |
| Qwen3.5-35B-A3B | Qwen | 0.545 | 2026-02-24 | evidence |
| Qwen3.6 Plus | Qwen | 0.625 | 2026-04-02 | evidence |
| Qwen3.7-Plus | Qwen | 0.733 | 2026-05-31 | evidence |
| Qwen3.8 Max | Qwen | 0.861 | 2026-08-02 | evidence |
| Qwen3.8-27B | Qwen | 0.843 | 2026-08-14 | evidence |
Scores are partitioned by the source that reported them and are never merged into a single cross-source ranking, because the sources measure different things and say so.