MetaMuse Spark 1.1
Best configuration ranks #29 of 610 on the BenchLeader Index at 64.9 ±3.8. Last measured 8 Sept 2026. Released 9 Jul 2026.
- Blended price
- $2.00/M
- $1.25 in · $4.25 out
- Output speed
- 194 tok/s
- First answer
- 2.03 s
- Context
- 1.0M
- Overall index65
- Reasoning69
- Coding68
- Agents & tools63
- Maths52
- Knowledge73
- Instruction following78
- Human preference66
- Multimodal65
- Long context65
- Composite72
Reasoning-effort configurations
The same model behaves differently depending on how much it is allowed to think. Each row is one setting, scored only on the benchmarks that were run at that setting. “Default” means the publisher did not say which setting was used.
| Effort | Index | Rank | Speed | First answer | Chat reply cost | Categories |
|---|---|---|---|---|---|---|
| xhigh | 56.4 | #158 | 194 tok/s | 2.03 s | $0.0018 | Agents & tools 58 · Coding 62 · Composite 52 · Knowledge 65 |
| defaultbest | 64.9 | #29 | 194 tok/s | 2.03 s | $0.0018 | Agents & tools 63 · Coding 68 · Composite 72 · Human preference 66 · Instruction following 78 · Knowledge 73 · Long context 65 · Maths 52 · Multimodal 65 · Reasoning 69 |
Benchmark results
One column per reasoning effort. Rank is among every configuration of every model on that benchmark. Hover a score for the run it came from.
Reasoning
| Benchmark | xhigh | default | Source | Trend |
|---|---|---|---|---|
| LMArena Hard Prompts | – | 1511#13 | LMArena | |
| LiveBench Reasoningnot in index | 87.7%#19 | – | LiveBench | |
| GPQA Diamond (AA)not in index | – | 89.8%#68 | Artificial Analysis | |
| Humanity's Last Exam (AA)not in index | – | 46.2%#25 | Artificial Analysis | |
| GPQA Diamond (Vals)not in index | 91.2%#23 | – | Vals AI |
Coding
| Benchmark | xhigh | default | Source | Trend |
|---|---|---|---|---|
| SciCode | – | 58.2%#7 | SciCode | |
| LMArena Coding | – | 1531#12 | LMArena | |
| LMArena WebDev | – | 1541#26 | LMArena | |
| LiveBench Codingnot in index | 77.2%#28 | – | LiveBench | |
| SciCode (AA)not in index | – | 58.8%#8 | Artificial Analysis | |
| LiveCodeBench | 85.9%#27 | – | Vals AI | |
| SWE-bench (Vals)not in index | 82.0%#21 | – | Vals AI | |
| SWE-Bench Pro | – | 61.5%#1 | Scale AI SEAL |
Agents & tools
| Benchmark | xhigh | default | Source | Trend |
|---|---|---|---|---|
| APEX-Agents | – | 41.9%#7 | Mercor | |
| LMArena Agent | – | -2.6#29 | LMArena | |
| LiveBench Agentic Codingnot in index | 58.5%#14 | – | LiveBench | |
| Terminal-Bench 2.1 (Vals) | 69.3%#23 | – | Vals AI | |
| MCP Atlas | – | 88.1%#1 | Scale AI SEAL |
Maths
| Benchmark | xhigh | default | Source | Trend |
|---|---|---|---|---|
| ProofBench | – | 39.0%#24 | Vals AI | |
| LiveBench Mathematicsnot in index | 87.1%#34 | – | LiveBench |
Knowledge
| Benchmark | xhigh | default | Source | Trend |
|---|---|---|---|---|
| SimpleQA Verified | – | 57.8%#14 | Epoch AI Benchmarking Hub | |
| LiveBench Data Analysisnot in index | 72.5%#33 | – | LiveBench | |
| AA-Omniscience | – | 28.1#23 | Artificial Analysis | |
| MMLU-Pro | 88.7%#17 | – | Vals AI | |
| LegalBench | 85.0%#24 | – | Vals AI | |
| CorpFin | 71.3%#4 | – | Vals AI | |
| TaxEval | 79.7%#2 | – | Vals AI | |
| PRBench Finance | – | 55.0%#1 | Scale AI SEAL | |
| PRBench Legal | – | 57.0%#1 | Scale AI SEAL | |
| MultiNRC | – | 65.6%#1 | Scale AI SEAL |
Instruction following
| Benchmark | xhigh | default | Source | Trend |
|---|---|---|---|---|
| LiveBench Languagenot in index | 74.3%#42 | – | LiveBench | |
| MultiChallenge | – | 75.3%#2 | Scale AI SEAL |
Human preference
| Benchmark | xhigh | default | Source | Trend |
|---|---|---|---|---|
| LMArena Text | – | 1492#10 | LMArena | |
| EQ-Bench 4 | – | 1260#8 | EQ-Bench |
Multimodal
| Benchmark | xhigh | default | Source | Trend |
|---|---|---|---|---|
| LMArena Vision | – | 1293#21 | LMArena |
Long context
| Benchmark | xhigh | default | Source | Trend |
|---|---|---|---|---|
| AA-LCR | – | 77.7%#87 | Artificial Analysis |
Composite
| Benchmark | xhigh | default | Source | Trend |
|---|---|---|---|---|
| Epoch Capabilities Indexnot in index | – | 154.6#27 | Epoch AI Benchmarking Hub | |
| LiveBench | 75.3%#27 | – | LiveBench | |
| AA Intelligence Index | – | 34.3#59 | Artificial Analysis | |
| Vals Indexnot in index | 54.8#21 | – | Vals AI |
Where it wins
Benchmarks where this configuration ranks in the top five of every configuration measured.
Where to run it
Every provider serving this model through OpenRouter, with throughput and first-token latency measured on live traffic over the last 30 minutes and each provider’s own price. Purple marks the best in each column.
| Provider | Speed | First token | Input $/M | Output $/M | Context | Quantisation |
|---|---|---|---|---|---|---|
| Meta | 194 tok/s | 2.03 s | $1.25 | $4.25 | 1.0M | – |
What a task costs
Estimates from list price, output speed and time to first answer for the best configuration. “With caching” assumes three-quarters of the input is served from the prompt cache at $0.150 per 1M. Reasoning tokens are not modelled.
| Workload | Tokens in / out | Cost | With caching | Time |
|---|---|---|---|---|
| Chat reply | 400 / 300 | $0.0018 | $0.0014 | 3.6 s |
| Summarise a 30-page report | 12,000 / 600 | $0.018 | $0.0077 | 5.1 s |
| Code edit | 6,000 / 1,500 | $0.014 | $0.0089 | 9.8 s |
| Agentic coding session | 60,000 / 4,000 | $0.092 | $0.043 | 22.7 s |
| Structured extraction | 2,000 / 200 | $0.0034 | $0.0017 | 3.1 s |
See also
Data as of 9 Sept 2026. Compare these configurations.