Meta Releases Muse Spark 1.1 with API Access and Enhanced Tool Calling
Meta has released Muse Spark 1.1, the first model in the Spark series to offer API access. The company claims significant improvements in agentic tool calling and computer use compared to the original Muse Spark released in April 2026.
Muse Spark 1.1 — Quick Specs
Meta Releases Muse Spark 1.1 with API Access and Enhanced Tool Calling
Meta has released Muse Spark 1.1, the first model in the Spark series to offer API access. The company claims significant improvements in agentic tool calling and computer use compared to the original Muse Spark released in April 2026.
The release marks a shift from the demo-only access of the original Muse Spark to a production-ready API that developers can integrate into applications.
Technical Details
Meta published a Muse Spark 1.1 Evaluation Report with additional technical details, though specific benchmark scores and pricing information have not yet been disclosed.
According to Meta, the model shows particular strength in:
- Agentic tool calling capabilities
- Computer use tasks
- SVG generation (demonstrated in early testing)
Developer Access
The model is accessible through Meta's API and can be used via the new llm-meta-ai plugin for the LLM command-line tool. Installation requires:
uv tool install llm
llm install llm-meta-ai
llm keys set meta-ai
llm -m meta-ai/muse-spark-1.1 "your prompt"
Early testing shows the model can generate SVG graphics, with one example successfully producing a recognizable pelican on a bicycle, though with somewhat blocky rendering.
What This Means
Muse Spark 1.1 represents Meta's move from experimental model releases to production-ready APIs. The focus on tool calling and computer use suggests Meta is positioning Spark as an agentic model to compete with Claude's computer use capabilities and OpenAI's function calling features. However, without published benchmark scores or pricing details, it's difficult to assess how Muse Spark 1.1 compares to existing models in the category. The release of an official API and developer tools indicates Meta is serious about this model line, following the pattern set by Llama of moving from research releases to production deployments.
Related Articles
OpenAI Halts Parts of Astra Model Development After It Hit 'Critical' Cybersecurity Threshold
OpenAI disclosed that its in-development Astra model showed cyberattack capabilities strong enough that it cannot rule out a 'Critical' risk classification. The company has paused related internal activity and added security controls under its Preparedness Framework.
Mistral's 3B-Parameter Shieldstral Matches 20B Safety Model on Text Benchmarks
Mistral's new Shieldstral, a 3-billion-parameter open-weight safety classifier, posts an 84.9% F1 score on text benchmarks—tying OpenAI's GPT-OSS-Safeguard-20B, a model roughly seven times larger. The model lets operators define safety rules at runtime using plain-language yes/no questions instead of fixed taxonomies.
Mistral AI Releases Shieldstral-1.0-3B, a 3B-Parameter Policy-Adaptive Safety Classifier
Mistral AI has released Shieldstral-1.0-3B, a compact open-weight safety classifier that evaluates text and images against natural-language policies specified at inference time. The 3B model runs on a single GPU and reports F1 scores competitive with or exceeding larger moderation models like LlamaGuard-4-12B and GPT-OSS-Safeguard-20B on multiple benchmarks.
Black Forest Labs Launches FLUX 3 Video, Claims It Beats Seedance 2.0 on Elo Rankings
Black Forest Labs has made FLUX 3 Video generally available via its API, offering up to 20-second HD/Full HD clips with native audio and lip-sync in 14+ languages. The company claims its internal Elo benchmarks put the model ahead of Seedance 2.0, Gemini Omni Flash, and Minimax H3.
Comments
Loading...