PLATFORMS Signal 94
AI Gateway adds Meta’s Muse Spark 1.3 with 1M-token context and multi-format input
↗
Vercel’s AI Gateway now hosts Muse Spark 1.3, a model from Meta with a 1M-token context window and support for text, images, and PDFs.
Engineers building or maintaining AI-powered agents and coding tools can now access a model with a larger context window and multi-format input. The contributor pricing tier offers a lower-cost option but trades off data privacy for training purposes.
Written by elseif from the cluster below · every claim links back to a sourceThe three things worth knowing
Muse Spark 1.3 improves agent workflows and coding tasks with fewer turns and less filler than prior versions
The model supports a 1M-token context window and accepts text, images, and PDFs as input
Two pricing tiers are available: standard and contributor, with the latter using inputs for training at a lower cost
THE READ
What the cluster adds up to.
Muse Spark 1.3 is now accessible via Vercel’s AI Gateway, marking an update to Meta’s model lineup. The model retains the same pricing structure as its predecessor, Muse Spark 1.2, but introduces a larger context window and expanded input capabilities. This change allows engineers to process longer prompts and incorporate non-textual data like images and PDFs directly into workflows.
The model’s improvements in agent work and coding are framed as reducing the number of turns required to complete tasks and minimizing filler output. This could streamline interactions for developers using the model in coding agents or automated workflows. However, the material does not specify the exact benchmarks or scenarios where these improvements are most pronounced, leaving adoption decisions dependent on testing in real-world use cases.
Two pricing tiers are available: standard and contributor. The standard tier operates under typical pricing, while the contributor tier offers significantly lower costs in exchange for Meta using the inputs and outputs for training. This trade-off may appeal to cost-sensitive projects but could be a non-starter for applications handling sensitive or proprietary data. The contributor tier’s pricing is nearly an order of magnitude lower, which may offset privacy concerns for some teams.
Integration with AI Gateway simplifies access for developers already using the platform. The model can be invoked via a straightforward API call, and the platform supports connections to coding agents like Claude Code and Codex. This reduces the overhead of adopting the model but ties its usage to Vercel’s ecosystem. Teams outside this ecosystem may need to evaluate whether the benefits outweigh the costs of integration or migration.
Written by elseif from the cluster below · checked for specifics the sources never containedTHE CLUSTER