For developers running long coding agents, the Contributor tier is remarkably cheap; Meta lists no countries for the API.
Kenning's take
What it does
Muse Spark 1.3 is the latest Muse Spark model, tuned for long-running agent work and coding. Meta says it asks clarifying questions when a prompt is unclear, confirms before consequential actions, follows long instructions more reliably and handles several tasks in one thread. It is available in Muse Code and the Meta Model API, including a max reasoning level, with a 1 million token context window.
- Max reasoning effort level
- 1 million token context window
- Asks clarifying questions and confirms risky actions
- Better multitasking in one long thread
- Cheaper Contributor tier that allows training on prompts
Who it suits. Developers building coding and agent tools on Meta's models.
Key facts
- Category
- AI models
- Part of
- Meta AI
- Made by
- Meta (Meta Superintelligence Labs)
- Released
- 2 September 2026
- How to get it
- in Muse Code and the Meta Model API as muse-spark-1.3 (Standard) or muse-spark-1.3-contributor (discounted)
- Works on
- API, CLI
- API price
- $1.25 per million input tokens, $4.25 per million output, $0.15 cached input (Standard tier); Contributor tier $0.10 input and $0.20 output in exchange for allowing training on prompts
FAQ
When was Muse Spark 1.3 released?
Meta released it on 2 September 2026.
How much does Muse Spark 1.3 cost?
$1.25 per million input tokens and $4.25 per million output on the Standard tier, or $0.10 and $0.20 on the Contributor tier, which lets Meta train on prompts.
Where can I use Muse Spark 1.3?
In Muse Code and through the Meta Model API at dev.meta.ai.
Facts from the maker's own pages and the sources listed. Checked 2026-10-11.
Similar tools
Meta Llama models
Meta's family of downloadable large language models, from Llama 2 to the multimodal Llama 4
Muse Glimmer
Meta's open-weight 30B model for running AI agents locally on a Mac or PC with one consumer GPU
Muse Image
Meta's in-house image model that plans a layout before drawing, used in Meta AI and the Meta Model API
Muse Spark
The first model from Meta Superintelligence Labs, a multimodal reasoning model behind the Meta AI app