Meta
availableShows if the model has enough results for an index.Muse Spark 1.1
Muse Spark 1.1 is a reasoning model from Meta in the Muse Spark family. 40 benchmarks count toward its score, in 7 categories.
IndexOverall score. 50 is the middle.67.0 ±3.1
CoverageShare of the index weight with results.95%
SpeedOutput tokens per second.217/s
Input / 1MUS dollars per 1M input tokens.$1.25
Output / 1MUS dollars per 1M output tokens.$4.25
ContextMaximum tokens in one request.1.05M
EloLMArena rating and rank.1480 (#15)
50 is the middle of the board. The range shows the doubt in the index.
27,615 votes. Elo shows what people prefer. It does not change the score.
CapabilitiesScore per category. 50 is the middle.
50 is the middleResults
40 counted| BenchmarkThe test name. | CategoryThe capability that the test measures. | ResultThe score from the publisher. | IndexThis result on the index scale. | RunThe settings of the run. | DateDate of the result. | Published byThe source of the result. |
|---|---|---|---|---|---|---|
| Cybench | Agentic | 92.9% | — | — | — | Andy K. Zhang et al. |
| GPQA Diamond | Knowledge | 91.2% | 62.4 | — | 1 Sept 2026 | Vals AI |
| Artificial Analysis GPQA Diamond | Knowledge | 89.8% | 60.6 | — | — | Artificial Analysis |
| MMLU Pro | Knowledge | 88.7% | 60.3 | — | 1 Sept 2026 | Vals AI |
| CharXiv Reasoning | Multimodal | 88.4% | 66.8 | — | — | CharXiv authors |
| MCP Atlas | Agentic | 88.1% | 74.4 | — | — | OpenAI |
| LiveBench Reasoning | Reasoning | 87.7% | 77.8 | xhigh effort | 25 Jun 2026 | LiveBench |
| LiveBench Mathematics | Math | 87.1% | 64.3 | xhigh effort | 25 Jun 2026 | LiveBench |
| MMMU Pro | Multimodal | 86.6% | 68.3 | — | 1 Sept 2026 | Vals AI |
| LiveCodeBench | Coding | 85.9% | 62.4 | — | 1 Sept 2026 | Vals AI |
| DeepSearchQA | Agentic | 84.9% | 65.5 | — | — | Meta AI |
| SWE-bench | Coding | 82.0% | 63.3 | — | 1 Sept 2026 | Vals AI |
| OSWorld-Verified | Agentic | 80.8% | 69.0 | — | — | Tianbao Xie et al. |
| Artificial Analysis Long Context Reasoning | Reasoning | 77.7% | 62.0 | — | — | Artificial Analysis |
| LiveBench Coding | Coding | 77.2% | 65.8 | xhigh effort | 25 Jun 2026 | LiveBench |
| BabyVision | Multimodal | 76.3% | — | — | — | Meta AI |
| Toolathlon | Agentic | 75.6% | 88.5 | — | — | OpenAI |
| LiveBench Language | Knowledge | 74.3% | 60.3 | xhigh effort | 25 Jun 2026 | LiveBench |
| LiveBench Data Analysis | Reasoning | 72.5% | 56.7 | xhigh effort | 25 Jun 2026 | LiveBench |
| Vibe Code Bench v1.1 | Coding | 72.2% | 72.2 | OpenHands | 21 Sept 2026 | Vals AI |
| Artificial Analysis Coding Index | Coding | 71.3% | 69.2 | — | — | Artificial Analysis |
| LiveBench Instruction Following | Instruction | 69.6% | 69.9 | xhigh effort | 25 Jun 2026 | LiveBench |
| Terminal-Bench 2.1 | Agentic | 69.3% | 64.9 | — | 21 Sept 2026 | Vals AI |
| WebArena-Verified Browser Agent Benchmark | Agentic | 69.0% | — | — | — | Amine El Hattami et al. |
| Humanity's Last Exam | Knowledge | 62.1% | 81.4 | — | — | Center for AI Safety et al. |
| SWE-bench Pro | Coding | 61.5% | 63.5 | — | — | Xiang Deng et al. |
| HealthBench Professional | Knowledge | 59.3% | — | — | — | Rebecca Soskin Hicks et al. |
| SkillsBench | Coding | 59.2% | 73.2 | OpenHands | 11 Sept 2026 | Vals AI |
| CyberGym | Agentic | 59.0% | 56.6 | — | — | Zhun Wang et al. |
| Artificial Analysis SciCode | Coding | 58.8% | 74.5 | — | — | Artificial Analysis |
| LiveBench Agentic Coding | Agentic | 58.5% | 71.8 | xhigh effort | 25 Jun 2026 | LiveBench |
| SimpleQA Verified | Knowledge | 57.8% | 75.1 | — | — | Epoch AI |
| JobBench | Agentic | 54.7% | 71.7 | — | — | Yuetai Li et al. |
| MRCR 1M | Reasoning | 54.1% | — | — | — | DeepSeek-AI |
| DeepSWE | Agentic | 53.3% | 62.5 | — | — | Datacurve AI |
| Humanity's Last Exam without tools | Knowledge | 52.2% | 73.0 | — | — | OpenAI |
| Artificial Analysis Omniscience Accuracy | Knowledge | 52.1% | 78.3 | — | — | Artificial Analysis |
| Artificial Analysis Humanity's Last Exam | Knowledge | 46.2% | 74.6 | — | — | Artificial Analysis |
| τ²-bench Banking | Agentic | 40.5% | 27.9 | xhigh effort · Sierra | 4 Aug 2026 | Sierra Research |
| GDPval-AA normalized | Agentic | 35.4% | 62.8 | — | — | Artificial Analysis |
| Artificial Analysis Intelligence Index | Knowledge | 33.7% | 64.6 | — | — | Artificial Analysis |
| Code Migration | Coding | 31.1% | 64.8 | — | 21 Sept 2026 | Vals AI |
| Artificial Analysis Agentic Index | Agentic | 27.5% | 59.6 | — | — | Artificial Analysis |
| Critical Physics Tasks | Reasoning | 15.1% | 70.0 | — | — | Artificial Analysis |
| OSWorld 2.0 | Agentic | 14.2% | 62.7 | — | — | Mengqi Yuan et al. |
| ExploitGym | Agentic | 0.8% | 61.9 | — | — | Zhun Wang et al. |
| Agent Arena task outcome | Agentic | 0.0 | 67.4 | — | 15 Sept 2026 | LMArena |
| ProgramBench | Coding | 0.0% | — | — | 21 Sept 2026 | Vals AI |
| Agent Arena command recovery | Agentic | -0.6 | 66.7 | — | 15 Sept 2026 | LMArena |
| Agent Arena steerability | Agentic | -4.2 | 62.6 | — | 15 Sept 2026 | LMArena |
40 benchmarks count, from 44 of 50 results. A grey row does not count. Too few models took that benchmark.