Muse Spark 1.2 is now available on Vercel AI Gateway
Meta's Muse Spark 1.2 lands on Vercel AI Gateway with coding-focused upgrades for long-horizon development workflows.
Why it matters
A coding-focused model update gains distribution through a unified API platform. Practitioners building with AI agents for code generation and repository-scale work now have another option integrated into a gateway that tracks cost and enables failover — modest but real optionality in the dev-tools stack.
The key facts
14 to knowMeta Muse Spark 1.2 released
Improvements: code generation, complex debugging, codebase understanding, end-to-end developer workflows
Supports long-horizon work: whole repository generation, large project builds, iterative improvement loops
Available via Vercel AI Gateway with model ID: meta/muse-spark-1.2
Works with coding agents (Claude Code, Codex, OpenCode, Pi)
AI Gateway features: custom reporting, Zero Data Retention, API key budgets, routing rules, failover
No platform fee on inference; provider pricing reflected with no markup
BYOK (Bring Your Own Key) support included
Muse Spark 1.2 ships with improvements in code generation, debugging, codebase understanding, and end-to-end developer workflows
Built for long-horizon work: generating whole repositories, multi-round iterative loops (write, compile, profile, improve)
Available via Vercel AI Gateway with unified API, usage tracking, cost reporting, retries, failover, and routing rules
Integrates with Claude Code, Codex, OpenCode, Pi in coding-agent mode
AI Gateway reflects provider pricing with no markup and no platform fee on inference
Published August 5, 2026
Go to the source
Vercel Blogvercel.com
Publisher excerpt: from Meta is now available on AI Gateway. It is a coding-focused update to the previous Muse Spark model. While keeping its general capabilities, 1.2 ships with improvements in code generation, complex debugging, codebase understanding, and end-to-end developer workflows.Muse Spark 1.2 The model is…