Meta Muse Spark 1.1

Metatextimageaudiovideo

Announced July 9, 2026 as the second model from Meta Superintelligence Labs and the launch model for the Meta Model API, marking the end of Meta's open-weights-only era and its first paid API. A multimodal reasoning model built for agentic work: it accepts text, images, video, audio and PDFs, returns text, and is designed to orchestrate multi-agent workflows as either a planning main agent or a subagent. The API is OpenAI-compatible with structured output and parallel tool calling, and new accounts get $20 in free credits. Mark Zuckerberg pitched the pricing as roughly a quarter of what Anthropic and OpenAI charge for comparable models.

Meta Muse Spark 1.1 strengths

  • Aggressive pricing for an agentic frontier-tier model
  • Broad input support: text, image, video, audio, PDF
  • 1M-token context window
  • Built for multi-agent orchestration (main agent or subagent)
  • OpenAI-compatible API with parallel tool calling

Pricing & context

Context window1,048,576 tokens
Input price /1M$1.25
Output price /1M$4.25
Modalitiestext, image, audio, video

Cost guide: a typical call of about 10K input + 2K output tokens costs roughly $0.021 at list prices. Worth modelling against cheaper tiers before committing high-volume traffic.

When to choose Meta Muse Spark 1.1

Meta Muse Spark 1.1 is best for cost-sensitive coding and computer-use agents that need long context and a drop-in OpenAI-compatible API. If your workload is more cost-sensitive, weigh it against gpt-oss-120b (≈$0.03 input /1M) first.

Meta Muse Spark 1.1 FAQ

How much does Meta Muse Spark 1.1 cost?

Meta Muse Spark 1.1 is priced at $1.25 per 1M input tokens and $4.25 per 1M output tokens (public API list price), with a 1,048,576 tokens context window. A typical call of about 10K input and 2K output tokens costs roughly $0.021.

What is Meta Muse Spark 1.1 best for?

Meta Muse Spark 1.1 by Meta is best for cost-sensitive coding and computer-use agents that need long context and a drop-in OpenAI-compatible API.

How does Meta Muse Spark 1.1 pricing compare to DeepSeek V4 Pro?

Meta Muse Spark 1.1 input costs $1.25 per 1M tokens versus $0.435 for DeepSeek V4 Pro, roughly 2.9x more expensive on input. Output is $4.25 vs $0.87.

Is Meta Muse Spark 1.1 multimodal?

Meta Muse Spark 1.1 supports text, image, audio, video.

Tools that use Meta Muse Spark 1.1

Other models

All models →
01Claude Fable 5Anthropic$10.00
02Claude Opus 5Anthropic$5.00
03GPT-5.6 SolOpenAI$5.00
04GPT-5.5OpenAI$5.00