New reasoning effort param for harder problems and frontier performance. Run /model, select Muse Spark 1.3. Then choose max for reasoning effort
Muse Spark 1.3 at a glance
Multimodal reasoning for long-horizon agentic and coding workflows.
| Spec | Muse Spark 1.3 |
|---|---|
| Model ID | meta/muse-spark-1.3 |
| Context window | 1.05M tokens |
| Released | September 2, 2026 |
| Intelligence Index | 48.1 |
| Coding Index | 75.8 |
| Input | $1.25 per 1M tokens |
| Output | $4.25 per 1M tokens |
| Cache read | $0.15 per 1M tokens |
| Effective input in an agent loop | $0.48 per 1M tokens |
| Plans | GOAT and above |
Specs, pricing and plans are from the Muse Spark 1.3 model page as of September 28, 2026, including any deal running that day, so they can differ from the launch pricing above. Intelligence and Coding Index scores are from Artificial Analysis.
Use Muse Spark 1.3
1cmd --model meta/muse-spark-1.3Or type /model in a session and pick it. You can switch mid-session without losing context.
What real work costs on Muse Spark 1.3
| Task | Tokens (in · out) | Cost |
|---|---|---|
| Quick lookup / one-liner | 8K · 1K | $0.0063 |
| Review a 500-line PR | 60K · 4K | $0.05 |
| Fix a bug (agent loop) | 180K · 12K | $0.13 |
| Refactor a module | 320K · 20K | $0.23 |
| Full-repo agent run | 900K · 45K | $0.54 |
Try Command Code
The best coding agent for open models, in your terminal or on your desktop.
1npm i -g command-codeDesktop App · Docs · X · Discord
