Muse Spark
Meta's mid line — and its open-weight release.
Muse Spark is the middle of Meta's lineup, with Muse Spark Open (816B) as the open-weight release — the most capable US open-weight model, in Meta's own words, served on SGLang, vLLM and Apple MLX. Meta publishes no prices or benchmarks for the Spark line.
What it is good at
- Spark Open 816B: open weights, run it yourself on SGLang, vLLM or Apple MLX
- The coding-optimized 1.2 is also served through OpenRouter
What it is less good at
- No prices, context windows or dates published on Meta's developer pages
- No TeamCLI route
Reaching it with TeamCLI
TeamCLI cannot reach Muse Spark yet — no verified or listed route carries it. It is listed because Meta offers it today.
Versions
One block per version, the current one first, each with its own anchor. The comparison table puts the costs and the family's benchmark columns side by side; the values are verbatim, and every one is sourced and dated in its version's section below and in the Sources list. A dash means the source publishes no figure, and nothing here is estimated.
No independent intelligence or speed index is republished here: Artificial Analysis measures both across labs, and its terms allow the link but not the copy — its figures live on its own pages.
| Version | Released | Context | Input price | Output price |
|---|---|---|---|---|
| Muse Spark 1.3 — current | — | — | — | — |
| Muse Spark 1.2 | — | — | — | — |
| Muse Spark 1.1 | — | — | — | — |
| Muse Spark Open | — | — | — | — |
Muse Spark 1.3 — current
- Model id: —
Abilities
- Released: —
- Context window: —
- Max output: —
- Output speed: —
- Latency: —
Costs
- Input price: —
- Output price: —
Muse Spark 1.2
- Model id: —
Abilities
- Released: —
- Context window: —
- Max output: —
- Output speed: —
- Latency: —
Costs
- Input price: —
- Output price: —
Muse Spark 1.1
- Model id: —
Abilities
- Released: —
- Context window: —
- Max output: —
- Output speed: —
- Latency: —
Costs
- Input price: —
- Output price: —
Muse Spark Open
- Model id: —
- 816B open-weight — the most capable US open-weight model to date, per Meta; SGLang, vLLM, Apple MLX
Abilities
- Released: —
- Context window: —
- Max output: —
- Output speed: —
- Latency: —
Costs
- Input price: —
- Output price: —
Sources
- Meta — developer model pages — read 2026-10-19
Figures refreshed 2026-10-19; benchmarks are vendor-reported. Prices are per million tokens unless a figure says otherwise.