Muse Spark 1.1
Muse Spark 1.1 is Meta's multimodal reasoning model for agentic work. Meta positions it for coding, computer use, tool orchestration, multimodal understanding, and long workflows through the public preview of the Meta Model API.
Muse Spark 1.1 combines model reasoning with native tool and computer-use ambitions, so readers need to evaluate more than benchmark claims. API access, region, tool compatibility, context behavior, price, and the difference between a hosted preview and downloadable weights all affect whether it fits a real agent workflow.
Meta's launch material controls model positioning, preview access, tool use, coding, multimodal behavior, context management, and orchestration claims. Community discussions are useful for questions about price, access, OpenCode support, and open-weight expectations, but they do not replace Meta documentation.
- Evaluate a Meta model for coding and agentic tool use.
- Build multimodal workflows that combine screenshots, text, and actions.
- Test computer-use and MCP-style tool orchestration in a hosted preview.
- Compare long-context agent behavior with other frontier model APIs.
Meta describes Muse Spark 1.1 as a multimodal reasoning model for agentic tasks. The launch emphasizes coding, computer use, tool calls, multimodal input, and workflows that need the model to plan and act rather than only answer a single prompt.
Meta says Muse Spark 1.1 can actively manage a one-million-token context and generalize across native tools, MCP servers, and custom skills. It also describes multi-agent orchestration, which makes the surrounding harness, permissions, traces, and evaluation plan as important as the model itself.
A public preview announcement is not a complete production contract. Before adopting Muse Spark 1.1, verify the current official model ID, regional access, pricing, rate limits, context limits, tool behavior, privacy terms, and provider support in the Meta Model API documentation and console.
Muse Spark 1.1 FAQ
Page-level questions for Muse Spark 1.1.
Is Muse Spark 1.1 open source or downloadable?+
Meta's launch presents Muse Spark 1.1 through the Meta Model API public preview. Do not assume the hosted model is downloadable or open weight unless Meta publishes a corresponding model card, repository, and license.
Does Muse Spark 1.1 support MCP tools?+
Meta says the model can generalize to native tools, MCP servers, and custom skills. Real compatibility still depends on the API surface and your harness, including tool schemas, permissions, retries, traces, and evaluation.
Should Muse Spark 1.1 be added to a production model router?+
Treat it as a candidate to test, not an automatic production default. Verify current access, region, price, model ID, rate limits, tool behavior, privacy terms, and task-specific evaluation results before routing production traffic.