GPT-5.6 Luna vs Muse Wonder Full
Full profiles: GPT-5.6 Luna (OpenAI) and Muse Wonder Full (Meta): every benchmark, price and rank.
Updated · Data from 9 sources
As of 25 Oct 2026, Muse Wonder Full (max) is the better pick for most workloads: it leads on quality (70.6 vs 59.9 on the BenchLeader Index). Muse Wonder Full leads in every category we measure both in.
GPT-5.6 Luna benchmarks 
Muse Wonder Full benchmarks Which to pick for…
The pick's numbers come first. Category scores on the BenchLeader scale (50 is the average model, 15 points one standard deviation). High confidence means a lead of half a standard deviation or more; under one point is called either.
| Use case | Pick | Why | Confidence |
|---|---|---|---|
| Coding | Muse Wonder Full | 72 vs 56 category score; IOI 79.8% vs 61.8% | High |
| Reasoning | Muse Wonder Full | 95 vs 63 category score; CritPt 32.9% vs 20.6% | High |
| Knowledge | Muse Wonder Full | 71 vs 57 category score; AA-Omniscience 30.1 vs -10.3 | High |
| Long documents | Either | 67 vs 67 category score; AA-LCR 83.7% vs 83.7% | Either |
| High volume on a budget | Neither flagship | A smaller model is usually the better fit: Cheapest good LLMs, Fastest LLMs | – |
Key differences
- Muse Wonder Full (max) leads on quality: 70.6 vs 59.9.
- GPT-5.6 Luna (max) is stronger in agents & tools, maths, multimodal.
- Muse Wonder Full (max) is stronger in coding, composite, knowledge, reasoning.
At a glance
| Spec | GPT-5.6 Luna | Muse Wonder Full |
|---|---|---|
| Developer | OpenAI | Meta |
| Released | 9 Jul 2026 | 12 Oct 2026 |
| Context window (tokens) | 1,050,000 | 1,000,000 |
| Compared at | GPT-5.6 Luna (max) | Muse Wonder Full (max) |
| Open weights | No | No |
| Price in / out / cached input per 1M | $0.200 / $1.20 / $0.020 | – |
Category by category
Category scores on the BenchLeader scale: 50 is the average evaluated model, 15 points is one standard deviation. A gap under one point is called level.
| Category | GPT-5.6 Luna (max) | Muse Wonder Full (max) | Leader |
|---|---|---|---|
| Reasoning | 62.9 | 94.9 | Muse Wonder Full |
| Coding | 56.0 | 72.1 | Muse Wonder Full |
| Agents & tools | 49.0 | – | not comparable |
| Maths | 65.0 | – | not comparable |
| Knowledge | 56.5 | 71.4 | Muse Wonder Full |
| Multimodal | 61.6 | – | not comparable |
| Long context | 67.0 | 67.0 | level |
| Composite | 71.2 | 91.9 | Muse Wonder Full |
Where sources disagree
Across the 33 benchmarks both have, GPT-5.6 Luna leads 0, Muse Wonder Full leads 31 and 2 are tied. Every overall index that rates both puts Muse Wonder Full ahead.
| Overall index | GPT-5.6 Luna (max) | Muse Wonder Full (max) | Leader |
|---|---|---|---|
| BenchLeader Index | 59.9 | 70.6 | Muse Wonder Full |
| AA Intelligence Index v4.3.2 | 37.3 | 55.8 | Muse Wonder Full |
Each publisher runs its own tasks, harnesses and effort settings, so their overall verdicts can differ; BenchLeader reconciles them on one scale. † measured at another effort setting of the same model (the publisher did not test the one compared here).
Price, speed and latency
List prices blended at three input tokens per output token. Per-task prices use the token counts on the methodology page; suite cost is what Artificial Analysis paid to run its whole Intelligence Index.
| Measure | GPT-5.6 Luna (max) | Muse Wonder Full (max) |
|---|---|---|
| Blended price per 1M tokens | $0.450 | – |
| Input / output per 1M | $0.200 / $1.20 | – / – |
| Cached input per 1M | $0.020 | – |
| Price per task: chat reply | $0.0004 | – |
| Price per task: summarize a 30-page report | $0.0031 | – |
| Price per task: code edit | $0.0030 | – |
| Price per task: agentic coding session | $0.017 | – |
| Price per task: structured extraction | $0.0006 | – |
| Cost to run a full benchmark suite | $0.178 | – |
| Output speed | 126 tok/s | – |
| Time to first answer | 62.0 s | – |
| Time to a complete answer | 65.9 s | – |
Each model is compared at its best-ranked reasoning-effort setting. GPT-5.6 Luna (max) scores 59.9 and takes 62 s to its first answer token; GPT-5.6 Luna (xhigh) scores 59.7 in 24 s, GPT-5.6 Luna (high) scores 57.0 in 7 s. Muse Wonder Full (max) scores 70.6. Higher effort means more thinking before the answer: better scores, longer waits and more tokens billed. How reasoning effort changes scores and cost.
Every benchmark
125 benchmarks, grouped by category; each name links to the board and its leaders. † measured at another effort setting of the same model, shown for reference and not counted as a win.
| Metric | Publisher | GPT-5.6 Luna (max) | Muse Wonder Full (max) |
|---|---|---|---|
| BenchLeader Index | BenchLeader | 59.9 | 70.6 |
| Agents & tools score | BenchLeader | 49.0 | – |
| Coding score | BenchLeader | 56.0 | 72.1 |
| Composite score | BenchLeader | 71.2 | 91.9 |
| Knowledge score | BenchLeader | 56.5 | 71.4 |
| Long context score | BenchLeader | 67.0 | 67.0 |
| Maths score | BenchLeader | 65.0 | – |
| Multimodal score | BenchLeader | 61.6 | – |
| Reasoning score | BenchLeader | 62.9 | 94.9 |
| Blended price $/M | BenchLeader | $0.450 | – |
| Output speed | BenchLeader | 126 tok/s | – |
| Time to first answer | BenchLeader | 62.0 s | – |
| Context window | BenchLeader | 1,050,000 tokens | 1,000,000 tokens |
| Reasoning | |||
| GPQA Diamond | Epoch AI Benchmarking Hub | 91.6% | – |
| SimpleBench | SimpleBench | 46.8%† | – |
| Arena Hard Prompts | Arena (LMArena) | 1475† | – |
| LiveBench 2.0 Reasoning | LiveBench | 85.6%† | – |
| LiveBench 1.0 Reasoning | LiveBench | 85.6% | – |
| GPQA Diamond (AA) | Artificial Analysis | 91.1% | – |
| Humanity's Last Exam (AA) | Artificial Analysis | 39.5% | 53.6% |
| GPQA Diamond (Vals) | Vals AI | 91.7% | – |
| Kagi LLM Benchmark | Kagi LLM Benchmark | 49.1%† | – |
| ARC-AGI-1 | ARC Prize | 90.7% | – |
| ARC-AGI-2 | ARC Prize | 59.6% | – |
| ARC-AGI-3 | ARC Prize | 0.2% | – |
| CritPt | Artificial Analysis | 20.6% | 32.9% |
| MysteryMechanism | Vals AI | 14.4% | 35.1% |
| Chess Puzzles | Epoch AI Benchmarking Hub | 43.0% | – |
| Mystery Game Puzzles | Epoch AI Benchmarking Hub | 21.0% | – |
| LMCA | Epoch AI Benchmarking Hub | 48.5% | – |
| DTBench | Epoch AI Benchmarking Hub | 88.8% | – |
| Coding | |||
| SciCode | SciCode | 53.6% | – |
| WeirdML | WeirdML | 60.9%† | – |
| FrontierCode | Cognition | 39.8%† | – |
| Arena Coding | Arena (LMArena) | 1502† | – |
| Arena WebDev | Arena (LMArena) | 1519† | 1726 |
| LiveBench 2.0 Coding | LiveBench | 73.6%† | – |
| LiveBench 1.0 Coding | LiveBench | 82.9% | – |
Show all 125 benchmarks
| Metric | Publisher | GPT-5.6 Luna (max) | Muse Wonder Full (max) |
|---|---|---|---|
| Coding | |||
| SciCode (AA) | Artificial Analysis | 53.6% | 62.6% |
| IOI | Vals AI | 61.8% | 79.8% |
| SWE-bench (Vals) | Vals AI | 93.0% | – |
| Code Migration | Vals AI | 44.5% | 61.0% |
| Code Migration v1.1 | Vals AI | – | 62.8% |
| ProgramBench | Vals AI | 0.0% | 0.5% |
| SWE-bench Multimodal | Vals AI | – | 47.3% |
| Vibe Code Bench 1-100 | Vals AI | 22.6% | 34.0% |
| Vibe Code Bench v1.1 | Vals AI | 77.1% | 86.0% |
| Surface Evolver Bench | Epoch AI Benchmarking Hub | 61.9%† | – |
| DeepSWE v1.1 | Epoch AI Benchmarking Hub | 67.2% | – |
| CursorBench | Cursor | 35.9% | – |
| ALE-Bench | Epoch AI Benchmarking Hub | 1667.4 | – |
| Terminal-Bench 4.0 (AA) | Artificial Analysis | 11.6% | 57.1% |
| Terminal-Bench 2.1 (AA) | Artificial Analysis | 80.9% | – |
| Startrise Frontend | Startrise | 65.0%† | – |
| FrontierCode 1.1 (AA) | Artificial Analysis | – | 44.9% |
| Agents & tools | |||
| Terminal-Bench | Terminal-Bench | 17.3% | – |
| APEX-Agents | Mercor | 43.0%† | – |
| LiveBench Long Agentic | LiveBench | 30.4%† | – |
| LiveBench Browser | LiveBench | 68.0%† | – |
| LiveBench Agentic Coding | LiveBench | 48.4% | – |
| Terminal-Bench 2.1 (Vals) | Vals AI | 79.0% | – |
| GDPval-AA v2.1 | Artificial Analysis | – | 67.8% |
| τ³-Banking (AA) | Artificial Analysis | 31.1% | – |
| ITBench SRE (AA) | Artificial Analysis | 40.3% | – |
| APEX-Agents (AA) | Artificial Analysis | 35.8% | – |
| CyberBench | Vals AI | 73.6% | 73.6% |
| Finance Agent v2 | Vals AI | 55.0% | 59.1% |
| Finance Agent v2.1 | Vals AI | – | 67.6% |
| Harvey's Legal Agent Benchmark v1.1 | Vals AI | – | 23.3% |
| Harvey's Legal Agent Benchmark v1 | Vals AI | 1.3% | 21.7% |
| Legal Research Bench | Vals AI | 36.5% | 53.4% |
| Robo-bench | Vals AI | – | 10.4% |
| SkillsBench | Vals AI | 60.5% | – |
| SREBench | Vals AI | 1.5% | 11.8% |
| Tax Agent Bench | Vals AI | 24.8% | 43.0% |
| Terminal-Bench 4.0 (Vals) | Vals AI | 11.6% | 58.1% |
| Terminal-Bench Science | Vals AI | 0.0% | 37.1% |
| BALROG | BALROG | 45.6% | – |
| Vending-Bench 2 | Andon Labs | 4094.7† | – |
| AutomationBench | Artificial Analysis | 50.2% | 71.3% |
| GDP.pdf (AA) | Artificial Analysis | 24.0% | 32.6% |
| Harvey LAB-AA v1.1 | Harvey | – | 13.1% |
| EnterpriseOps-Gym | Artificial Analysis | 40.8% | – |
| AA-Briefcase v1.1 | Artificial Analysis | 1339 | 1729 |
| Terminal-Bench-Science 0.1 (AA) | Artificial Analysis | 3.3% | 37.1% |
| FACTS Search | 79.3%† | – | |
| Maths | |||
| FrontierMath Tiers 1–3 | Epoch AI Benchmarking Hub | 82.1% | – |
| FrontierMath Tier 4 | Epoch AI Benchmarking Hub | 61.0% | – |
| OTIS Mock AIME | Epoch AI Benchmarking Hub | 98.3% | – |
| ProofBench | Vals AI | 60.0% | – |
| LiveBench Mathematics | LiveBench | 87.2% | – |
| Arena Maths | Arena (LMArena) | 1478† | – |
| Knowledge | |||
| SimpleQA Verified | Epoch AI Benchmarking Hub | 41.0% | – |
| LiveBench 2.0 Data Analysis | LiveBench | 78.0%† | – |
| LiveBench Research | LiveBench | 51.0%† | – |
| LiveBench 1.0 Data Analysis | LiveBench | 78.0% | – |
| AA-Omniscience | Artificial Analysis | -10.3 | 30.1 |
| MMLU-Pro | Vals AI | 86.0% | – |
| LegalBench | Vals AI | 84.0% | 87.4% |
| CorpFin | Vals AI | 64.2% | – |
| TaxEval | Vals AI | 76.2% | – |
| BioMysteryBench | Vals AI | 61.5% | 73.3% |
| Excel Modeling Benchmark | Vals AI | 67.1% | 71.9% |
| Excel Modeling Benchmark v1.1 | Vals AI | – | 75.0% |
| MedCode | Vals AI | 42.4% | 49.9% |
| MedScribe | Vals AI | 84.4% | 89.3% |
| Public Benefits Bench v1.1 | Vals AI | 61.2% | 73.8% |
| GDP.pdf | Epoch AI Benchmarking Hub | 22.7%† | – |
| AA-Omniscience: accuracy | Artificial Analysis | 42.7% | 43.2% |
| AA-Omniscience: non-hallucination | Artificial Analysis | 7.4% | 77.0% |
| Instruction following | |||
| LiveBench Language | LiveBench | 72.6% | – |
| LiveBench Instruction Following | LiveBench | 60.1% | – |
| Arena Instruction Following | Arena (LMArena) | 1444† | – |
| Human preference | |||
| Arena Text | Arena (LMArena) | 1453† | – |
| EQ-Bench 4 | EQ-Bench | 1149† | – |
| Arena Creative Writing | Arena (LMArena) | 1410† | – |
| Arena Multi-turn | Arena (LMArena) | 1459† | – |
| Arena Longer Queries | Arena (LMArena) | 1452† | – |
| Multimodal | |||
| Arena Vision | Arena (LMArena) | 1257† | – |
| LiveBench Design | LiveBench | 76.0%† | – |
| MMMU-Pro | Artificial Analysis | 78.5% | – |
| MMMU-Pro (Vals) | Vals AI | 85.0% | – |
| MortgageTax | Vals AI | 67.3% | – |
| SAGE | Vals AI | 44.2% | – |
| Blueprint-Bench 2 | Epoch AI Benchmarking Hub | 22.6%† | – |
| Long context | |||
| AA-LCR | Artificial Analysis | 83.7% | 83.7% |
| Arena Document | Arena (LMArena) | 1462† | – |
| MLCR | Artificial Analysis | 20.0% | – |
| Composite | |||
| Epoch Capabilities Index | Epoch AI Benchmarking Hub | 156.5† | – |
| LiveBench 2.0 | LiveBench | 66.1%† | – |
| LiveBench 1.0 | LiveBench | 73.6% | – |
| AA Intelligence Index v4.3.2 | Artificial Analysis | 37.3 | 55.8 |
| Vals Index | Vals AI | – | 66.8 |
| Vals Finance Index | Vals AI | – | 64.5% |
| Vals RSI Index | Vals AI | – | 24.4% |
| Vals Science Index | Vals AI | 25.3% | 48.5% |
| Vals Multimodal Index | Vals AI | 69.1% | – |
| Vals Router Index | Vals AI | – | 66.8%† |
Data as of 2026-10-25. Best configuration of each model; every score links to its source on the model pages.
What changed
- GPT-5.6 Luna (max): 1 new result (AA-Briefcase v1.1).
- Muse Wonder Full (max): 1 new result (Terminal-Bench-Science 0.1 (AA)).
- GPT-5.6 Luna (max): 2 new results (Harvey's Legal Agent Benchmark v1, Vals Science Index).
- Muse Wonder Full (max): 9 new results (Code Migration v1.1, Excel Modeling Benchmark v1.1, Finance Agent v2.1…).
- GPT-5.6 Luna (max): 1 score revision.
- Muse Wonder Full (max): 1 new result (FrontierCode 1.1 (AA)).
- Muse Wonder Full (max): BenchLeader Index 68.6 → 70.7.
- Muse Wonder Full (max): 1 new result (Arena WebDev); 2 score revisions.
- GPT-5.6 Luna (max): BenchLeader Index 58.8 → 59.9.
- GPT-5.6 Luna (max): 5 new results (AutomationBench, GDP.pdf (AA), MLCR…).
Related comparisons
- GPT-5.6 Luna vs Claude Opus 5.5
- GPT-5.6 Luna vs Gemini 4 Argon
- GPT-5.6 Luna vs Claude Fable 5.1
- Muse Wonder Full vs Claude Opus 5.5
- Muse Wonder Full vs Gemini 4 Argon
- Muse Wonder Full vs Claude Fable 5.1
- Muse Wonder Full vs GPT-6 Astra
- GPT-5.5 vs Muse Wonder Full
Rankings: Best LLM overall · Best LLM for coding · Best LLM for agents · Best cheap LLMs
GPT-5.6 Luna vs Muse Wonder Full: questions
- Is GPT-5.6 Luna better than Muse Wonder Full?
- Muse Wonder Full (max) leads on quality: 70.6 vs 59.9. The BenchLeader Index combines every independent quality benchmark; Muse Wonder Full (max) is ahead overall as of 2026-10-25, but check the category scores for your use.
- Is GPT-5.6 Luna better than Muse Wonder Full for coding?
- Muse Wonder Full scores higher in coding (72 vs 56 on the category index, where 50 is average). On IOI, the top shared coding benchmark, it is 61.8% for GPT-5.6 Luna against 79.8% for Muse Wonder Full.
- Which is better for long documents?
- They are level on long-context benchmarks (67 vs 67). Both take about 1.1M tokens: GPT-5.6 Luna 1,050,000, Muse Wonder Full 1,000,000.
- Do independent benchmarks agree on GPT-5.6 Luna vs Muse Wonder Full?
- Across the 33 benchmarks both have, GPT-5.6 Luna leads 0, Muse Wonder Full leads 31 and 2 are tied. Every overall index that rates both puts Muse Wonder Full ahead.
- When was Muse Wonder Full released?
- Muse Wonder Full was released on 12 Oct 2026.