Meta: Muse Spark 1.3 Contributor
Muse Spark 1.3 Contributor is the cost-efficient contributor tier of Meta’s multimodal reasoning model for experimentation, learning, and early-stage agentic, multi-agent, and coding workflows. It is designed to track information... Very low token pricing for a contributor-tier model 1M-token context for long-running workflows Accepts text, image, video, file, and audio inputs Contributor tier is below the full Muse Spark model New release has limited independent benchmark evidence Multimodal input does not imply multimodal output Early-stage agentic workflows Multimodal prototyping Budget coding and experimentation
Specifications
| Lab | Meta |
|---|---|
| Context window | 1,048,576 |
| Input price | $0.10/1M |
| Output price | $0.20/1M |
| Release | 2026-09-01 00:00:00 |
Strengths
- Very low token pricing for a contributor-tier model
- 1M-token context for long-running workflows
- Accepts text, image, video, file, and audio inputs
Weaknesses
- Contributor tier is below the full Muse Spark model
- New release has limited independent benchmark evidence
- Multimodal input does not imply multimodal output
Best for
- Early-stage agentic workflows
- Multimodal prototyping
- Budget coding and experimentation
In Depth: Muse Spark 1.3 Contributor
Summary
Muse Spark 1.3 Contributor is an AI model from Meta.
Released 2026-09-01 00:00:00. It supports Text, image, video, file, audio input and produces Text output, with a context window of 1M tokens. Input pricing is $0.10 per 1M tokens and output is $0.20 per 1M tokens on OpenRouter.