44 observations · 44 standings · first ingest Oct 2, 2026
Source record llm-stats-osworld · higher is better · declared max 1
| Subject | Value | Instrument | Protocol | Measured by | Reported | Document |
|---|---|---|---|---|---|---|
| Qwen3.8 Max | 0.86 | – | – | Qwen | Aug 2, 2026(model_announcement) | //api.zeroeval.com/leaderboard/benchmarks/osworld-verified?top_n=500 |
| Claude Fable 5 | 0.85 | – | – | Anthropic | Jun 9, 2026 |
| //api.zeroeval.com/leaderboard/benchmarks/osworld-verified?top_n=500 |
| Qwen3.8-27B | 0.84 | – | – | Qwen | Aug 14, 2026(model_announcement) | //api.zeroeval.com/leaderboard/benchmarks/osworld-verified?top_n=500 |
| Claude Opus 4.8 | 0.83 | – | – | Anthropic | May 28, 2026(model_announcement) | //api.zeroeval.com/leaderboard/benchmarks/osworld-verified?top_n=500 |
| Gemini 3.6 Flash | 0.83 | – | – | Jul 21, 2026(model_announcement) | //api.zeroeval.com/leaderboard/benchmarks/osworld-verified?top_n=500 |
| Claude Sonnet 5 | 0.81 | – | – | Anthropic | Jun 30, 2026(model_announcement) | //api.zeroeval.com/leaderboard/benchmarks/osworld-verified?top_n=500 |
| Muse Spark 1.1 | 0.81 | – | – | Meta | Jul 9, 2026(model_announcement) | //api.zeroeval.com/leaderboard/benchmarks/osworld-verified?top_n=500 |
| Claude Mythos Preview | 0.80 | – | – | Anthropic | Apr 7, 2026(model_announcement) | //api.zeroeval.com/leaderboard/benchmarks/osworld-verified?top_n=500 |
| Seed 2.1 Pro | 0.79 | – | – | ByteDance | Jun 24, 2026(model_announcement) | //api.zeroeval.com/leaderboard/benchmarks/osworld?top_n=500 |
| GPT-5.5 | 0.79 | – | – | OpenAI | Apr 23, 2026(model_announcement) | //api.zeroeval.com/leaderboard/benchmarks/osworld-verified?top_n=500 |
| Gemini 3.5 Flash | 0.78 | – | – | May 19, 2026(model_announcement) | //api.zeroeval.com/leaderboard/benchmarks/osworld-verified?top_n=500 |
| Claude Opus 4.7 | 0.78 | – | – | Anthropic | Apr 16, 2026(model_announcement) | //api.zeroeval.com/leaderboard/benchmarks/osworld-verified?top_n=500 |
| Seed 2.1 Turbo | 0.76 | – | – | ByteDance | Jun 24, 2026(model_announcement) | //api.zeroeval.com/leaderboard/benchmarks/osworld?top_n=500 |
| GPT-5.4 | 0.75 | – | – | OpenAI | Mar 5, 2026(model_announcement) | //api.zeroeval.com/leaderboard/benchmarks/osworld-verified?top_n=500 |
| Gemini 3.5 Flash-Lite | 0.74 | – | – | Jul 21, 2026(model_announcement) | //api.zeroeval.com/leaderboard/benchmarks/osworld-verified?top_n=500 |
| Qwen3.7-Plus | 0.73 | – | – | Qwen | May 31, 2026(model_announcement) | //api.zeroeval.com/leaderboard/benchmarks/osworld-verified?top_n=500 |
| Kimi K2.6 | 0.73 | – | – | Moonshot AI | Apr 20, 2026(model_announcement) | //api.zeroeval.com/leaderboard/benchmarks/osworld-verified?top_n=500 |
| Claude Opus 4.6 | 0.73 | – | – | Anthropic | Feb 5, 2026(model_announcement) | //api.zeroeval.com/leaderboard/benchmarks/osworld?top_n=500 |
| Claude Sonnet 4.6 | 0.72 | – | – | Anthropic | Feb 17, 2026(model_announcement) | //api.zeroeval.com/leaderboard/benchmarks/osworld?top_n=500 |
| GPT-5.4 mini | 0.72 | – | – | OpenAI | Mar 17, 2026(model_announcement) | //api.zeroeval.com/leaderboard/benchmarks/osworld-verified?top_n=500 |
| MiniMax M3 | 0.70 | – | – | MiniMax | Jun 1, 2026(model_announcement) | //api.zeroeval.com/leaderboard/benchmarks/osworld-verified?top_n=500 |
| Qwen3 VL 235B A22B Instruct | 0.67 | – | – | Qwen | Sep 22, 2025(model_announcement) | //api.zeroeval.com/leaderboard/benchmarks/osworld?top_n=500 |
| Claude Opus 4.5 | 0.66 | – | – | Anthropic | Nov 24, 2025(model_announcement) | //api.zeroeval.com/leaderboard/benchmarks/osworld?top_n=500 |
| Muse Glimmer-30B | 0.66 | – | – | Meta | Aug 10, 2026(model_announcement) | //api.zeroeval.com/leaderboard/benchmarks/osworld-verified?top_n=500 |
| GPT-5.3 Codex | 0.65 | – | – | OpenAI | Feb 5, 2026(model_announcement) | //api.zeroeval.com/leaderboard/benchmarks/osworld-verified?top_n=500 |
| Qwen3.6 Plus | 0.63 | – | – | Qwen | Apr 2, 2026(model_announcement) | //api.zeroeval.com/leaderboard/benchmarks/osworld-verified?top_n=500 |
| GLM-5V-Turbo | 0.62 | – | – | Z.ai | Apr 2, 2026(model_announcement) | //api.zeroeval.com/leaderboard/benchmarks/osworld?top_n=500 |
| Claude Sonnet 4.5 | 0.61 | – | – | Anthropic | Sep 29, 2025(model_announcement) | //api.zeroeval.com/leaderboard/benchmarks/osworld?top_n=500 |
| Qwen3.5-122B-A10B | 0.58 | – | – | Qwen | Feb 24, 2026(model_announcement) | //api.zeroeval.com/leaderboard/benchmarks/osworld-verified?top_n=500 |
| Qwen3.5-27B | 0.56 | – | – | Qwen | Feb 24, 2026(model_announcement) | //api.zeroeval.com/leaderboard/benchmarks/osworld-verified?top_n=500 |
| Qwen3.5-35B-A3B | 0.55 | – | – | Qwen | Feb 24, 2026(model_announcement) | //api.zeroeval.com/leaderboard/benchmarks/osworld-verified?top_n=500 |
| Claude Haiku 4.5 | 0.51 | – | – | Anthropic | Oct 15, 2025(model_announcement) | //api.zeroeval.com/leaderboard/benchmarks/osworld?top_n=500 |
| Qwen3 VL 32B Thinking | 0.41 | – | – | Qwen | Sep 22, 2025(model_announcement) | //api.zeroeval.com/leaderboard/benchmarks/osworld?top_n=500 |
| GPT-5.4 nano | 0.39 | – | – | OpenAI | Mar 17, 2026(model_announcement) | //api.zeroeval.com/leaderboard/benchmarks/osworld-verified?top_n=500 |
| Qwen3 VL 235B A22B Thinking | 0.38 | – | – | Qwen | Sep 22, 2025(model_announcement) | //api.zeroeval.com/leaderboard/benchmarks/osworld?top_n=500 |
| Qwen3 VL 8B Thinking | 0.34 | – | – | Qwen | Sep 22, 2025(model_announcement) | //api.zeroeval.com/leaderboard/benchmarks/osworld?top_n=500 |
| Qwen3 VL 8B Instruct | 0.34 | – | – | Qwen | Sep 22, 2025(model_announcement) | //api.zeroeval.com/leaderboard/benchmarks/osworld?top_n=500 |
| Qwen3 VL 32B Instruct | 0.33 | – | – | Qwen | Sep 22, 2025(model_announcement) | //api.zeroeval.com/leaderboard/benchmarks/osworld?top_n=500 |
| Qwen3 VL 4B Thinking | 0.31 | – | – | Qwen | Sep 22, 2025(model_announcement) | //api.zeroeval.com/leaderboard/benchmarks/osworld?top_n=500 |
| Qwen3 VL 30B A3B Thinking | 0.31 | – | – | Qwen | Sep 22, 2025(model_announcement) | //api.zeroeval.com/leaderboard/benchmarks/osworld?top_n=500 |
| Qwen3 VL 30B A3B Instruct | 0.30 | – | – | Qwen | Sep 22, 2025(model_announcement) | //api.zeroeval.com/leaderboard/benchmarks/osworld?top_n=500 |
| Qwen3 VL 4B Instruct | 0.26 | – | – | Qwen | Sep 22, 2025(model_announcement) | //api.zeroeval.com/leaderboard/benchmarks/osworld?top_n=500 |
| Qwen2.5 VL 72B Instruct | 0.09 | – | – | Qwen | Jan 26, 2025(model_announcement) | //api.zeroeval.com/leaderboard/benchmarks/osworld?top_n=500 |
| Qwen2.5 VL 32B Instruct | 0.06 | – | – | Qwen | Feb 28, 2025(model_announcement) | //api.zeroeval.com/leaderboard/benchmarks/osworld?top_n=500 |
| Subject | Value | Reported | Kind | Ingested |
|---|---|---|---|---|
| Qwen3.8-27B | 0.84 | Aug 14, 2026 | unspecified | Oct 2, 2026 |
| Muse Glimmer-30B | 0.66 | Aug 10, 2026 | unspecified | Oct 2, 2026 |
| Qwen3.8 Max | 0.86 | Aug 2, 2026 | unspecified | Oct 2, 2026 |
| Gemini 3.6 Flash | 0.83 | Jul 21, 2026 | unspecified | Oct 2, 2026 |
| Gemini 3.5 Flash-Lite | 0.74 | Jul 21, 2026 | unspecified | Oct 2, 2026 |
| Muse Spark 1.1 | 0.81 | Jul 9, 2026 | unspecified | Oct 2, 2026 |
| Claude Sonnet 5 | 0.81 | Jun 30, 2026 | unspecified | Oct 2, 2026 |
| Seed 2.1 Pro | 0.79 | Jun 24, 2026 | unspecified | Oct 2, 2026 |
| Seed 2.1 Turbo | 0.76 | Jun 24, 2026 | unspecified | Oct 2, 2026 |
| Claude Fable 5 | 0.85 | Jun 9, 2026 | unspecified | Oct 2, 2026 |
| MiniMax M3 | 0.70 | Jun 1, 2026 | unspecified | Oct 2, 2026 |
| Qwen3.7-Plus | 0.73 | May 31, 2026 | unspecified | Oct 2, 2026 |
| Claude Opus 4.8 | 0.83 | May 28, 2026 | unspecified | Oct 2, 2026 |
| Gemini 3.5 Flash | 0.78 | May 19, 2026 | unspecified | Oct 2, 2026 |
| GPT-5.5 | 0.79 | Apr 23, 2026 | unspecified | Oct 2, 2026 |
| Kimi K2.6 | 0.73 | Apr 20, 2026 | unspecified | Oct 2, 2026 |
| Claude Opus 4.7 | 0.78 | Apr 16, 2026 | unspecified | Oct 2, 2026 |
| Claude Mythos Preview | 0.80 | Apr 7, 2026 | unspecified | Oct 2, 2026 |
| GLM-5V-Turbo | 0.62 | Apr 2, 2026 | unspecified | Oct 2, 2026 |
| Qwen3.6 Plus | 0.63 | Apr 2, 2026 | unspecified | Oct 2, 2026 |
| GPT-5.4 mini | 0.72 | Mar 17, 2026 | unspecified | Oct 2, 2026 |
| GPT-5.4 nano | 0.39 | Mar 17, 2026 | unspecified | Oct 2, 2026 |
| GPT-5.4 | 0.75 | Mar 5, 2026 | unspecified | Oct 2, 2026 |
| Qwen3.5-122B-A10B | 0.58 | Feb 24, 2026 | unspecified | Oct 2, 2026 |
| Qwen3.5-27B | 0.56 | Feb 24, 2026 | unspecified | Oct 2, 2026 |
| Qwen3.5-35B-A3B | 0.55 | Feb 24, 2026 | unspecified | Oct 2, 2026 |
| Claude Sonnet 4.6 | 0.72 | Feb 17, 2026 | unspecified | Oct 2, 2026 |
| Claude Opus 4.6 | 0.73 | Feb 5, 2026 | unspecified | Oct 2, 2026 |
| GPT-5.3 Codex | 0.65 | Feb 5, 2026 | unspecified | Oct 2, 2026 |
| Claude Opus 4.5 | 0.66 | Nov 24, 2025 | unspecified | Oct 2, 2026 |
| Claude Haiku 4.5 | 0.51 | Oct 15, 2025 | unspecified | Oct 2, 2026 |
| Claude Sonnet 4.5 | 0.61 | Sep 29, 2025 | unspecified | Oct 2, 2026 |
| Qwen3 VL 235B A22B Instruct | 0.67 | Sep 22, 2025 | unspecified | Oct 2, 2026 |
| Qwen3 VL 32B Thinking | 0.41 | Sep 22, 2025 | unspecified | Oct 2, 2026 |
| Qwen3 VL 235B A22B Thinking | 0.38 | Sep 22, 2025 | unspecified | Oct 2, 2026 |
| Qwen3 VL 8B Thinking | 0.34 | Sep 22, 2025 | unspecified | Oct 2, 2026 |
| Qwen3 VL 8B Instruct | 0.34 | Sep 22, 2025 | unspecified | Oct 2, 2026 |
| Qwen3 VL 32B Instruct | 0.33 | Sep 22, 2025 | unspecified | Oct 2, 2026 |
| Qwen3 VL 4B Thinking | 0.31 | Sep 22, 2025 | unspecified | Oct 2, 2026 |
| Qwen3 VL 30B A3B Thinking | 0.31 | Sep 22, 2025 | unspecified | Oct 2, 2026 |
| Qwen3 VL 30B A3B Instruct | 0.30 | Sep 22, 2025 | unspecified | Oct 2, 2026 |
| Qwen3 VL 4B Instruct | 0.26 | Sep 22, 2025 | unspecified | Oct 2, 2026 |
| Qwen2.5 VL 32B Instruct | 0.06 | Feb 28, 2025 | unspecified | Oct 2, 2026 |
| Qwen2.5 VL 72B Instruct | 0.09 | Jan 26, 2025 | unspecified | Oct 2, 2026 |
4 observations · 4 standings · first ingest Oct 2, 2026
Source record osworld · higher is better · metric success
| Subject | Value | Instrument | Protocol | Measured by | Reported | Document |
|---|---|---|---|---|---|---|
| Kimi K3 | 84.8% | osworld_verified | reasoning=max | Moonshot AI | Jun 13, 2026(publication date) | moonshot_kimi_k3_model_card |
| Claude Opus 5 | 70.6% | osworld_2.0 | max effort, first-attempt success rate, averaged over 5 runs, 1080p, max 500 action steps | Anthropic | Jul 24, 2026(publication date) | anthropic_claude_opus_5_system_card |
| Qwen3.5-397B-A17B | 62.2% | osworld_verified | thinking (temp 0.6, top_p 0.95, top_k 20) | Qwen | Feb 16, 2026(publication date) | qwen3_5_model_card |
| Kimi K3 | 58.3% | osworld_2.0 | reasoning=max | Moonshot AI | Jun 13, 2026(publication date) | moonshot_kimi_k3_model_card |
| Subject | Value | Reported | Kind | Ingested |
|---|---|---|---|---|
| Claude Opus 5 | 70.6% | Jul 24, 2026 | self-reported | Oct 2, 2026 |
| Kimi K3 | 58.3% | Jun 13, 2026 | self-reported | Oct 2, 2026 |
| Kimi K3 | 84.8% | Jun 13, 2026 | self-reported | Oct 2, 2026 |
| Qwen3.5-397B-A17B | 62.2% | Feb 16, 2026 | self-reported | Oct 2, 2026 |
documents that define or report scores for this benchmark
| release post | Anthropic | Claude Fable 5 and Mythos 5Jun 9, 2026 |
| system card | Anthropic | Claude Opus 5Jul 24, 2026 |
| other | Moonshot AI | Kimi K3Jun 13, 2026 |
| other | Qwen | Qwen3.5-397B-A17BFeb 16, 2026 |
| other | Qwen | Qwen3.8-27BAug 5, 2026 |
| other | – | ubuntu_osworld– |
| other | – | OSWorld– |
| other | – | OSWorld-Verified– |
Observations and source documents are located via the tracking corpus. Every row cites its primary source document; observations from different sources stay in separate partitions and are never merged into a cross-source ranking.