Muse Spark 1.2 vs Muse Spark 1.3
| Benchmark | Muse Spark 1.2 | Muse Spark 1.3 |
|---|---|---|
| AutomationBench-AA | 40.6 | 56.8 |
| GDPval-AA (work tasks) 2.1 | 1,474.6 | 1,633.7 |
| τ³-Banking · AA | 34.8 | 47.2 |
| Terminal-Bench 4.0 | 7.1 | 16.7 |
| AA-Briefcase 1.1 · Elo | 1,326.9 | 1,482.3 |
| AA-LCR (long context) 1.1 | 79 | 83 |
| DeepSWE 1.1 | 54.9 | 67.8 · xhigh |
| Toolathlon-Verified | 75.9 | — |
| AA Analyst Agent | — | 37.5 · max |
| ITBench SRE | — | 33.2 · max |
| Terminal-Bench Science 0.1 | — | 3.8 |
| EnterpriseOps-Gym-AA | 47.3 | — |
| FrontierCode 1.1 | — | 38.6 |
| Harvey LAB-AA 1.1 | — | 8.9 · max |
Specs
| Spec | Muse Spark 1.2 | Muse Spark 1.3 |
|---|---|---|
| Input, per 1M | $1.25 | $1.25 |
| Output, per 1M | $4.25 | $4.25 |
| Cost per task? | $1.29 | $2.15 |
| Speed, tok/s | 292 | 301 |
| Context | 1M | 1M |
| Weights | Closed | Closed |
| Released | Aug 2026 | Sep 2026 |
| See images | ✓ | ✓ |
| Read PDFs | ✓ | ✓ |
| Hear audio | — | — |
| Watch video | ✓ | ✓ |
| Call tools | ✓ | ✓ |
| Force a tool call | — | — |
| Return strict JSON | ✓ | ✓ |
| Search the web | ✓ | ✓ |
| Run code | ? | ? |
| Use a computer | ? | ? |
What breaks when you switch
Only auto tool_choice is supported›
tool_choice none, required or a named function returns HTTP 400; use auto and handle either a tool call or text.
Search filters return HTTP 400 · Messages API›
allowed_domains, blocked_domains and max_uses return HTTP 400 on Messages search grounding; omit them and use the supported web_search tool.
Manual budget does not set effort · Messages API›
Compatibility budget_tokens does not set reasoning effort; use output_config.effort for depth control.
No Chat Completions search grounding · Chat Completions›
Search grounding is unavailable in Chat Completions; use Responses or Messages.
Reasoning cannot be disabled›
Muse Wonder and Muse Spark do not support disabling reasoning with none; keep reasoning enabled.
Private reasoning still costs tokens›
Private reasoning consumes the output budget and billed output tokens; budget for it even though it is absent from the response body.
Only auto tool_choice is supported›
tool_choice none, required or a named function returns HTTP 400; use auto and handle either a tool call or text.
Reasoning cannot be disabled›
Muse Wonder and Muse Spark do not support disabling reasoning with none; keep reasoning enabled.