Muse Wonder
Meta’s closed Muse Wonder (muse-wonder-1.0, 2026-10-12) on the Meta Model API and in Muse Code. Muse Spark 1.3 remains the lower-priced API.
Meta Superintelligence Labs’ Muse Wonder (API id muse-wonder-1.0, 2026-10-12) — closed model for longer agentic coding and professional tasks on the Meta Model API and in Muse Code. It sits above Muse Spark 1.3, which remains the lower-priced API. Muse Wonder Full is a Muse Code subscriber variant; the public pricing page lists only muse-wonder-1.0. Preview. Closed weights.
Why Muse Wonder matters
Wonder is Meta’s newer paid peak for longer agentic coding and professional tasks, above Spark 1.3. Standard price is $4 / $0.20 cached / $20 per 1M tokens, with no Contributor mode on the Model API. Preview, closed weights, 1,048,576-token context. Muse Wonder Full is for Muse Code subscribers and has no separate public token price.
Vision · Audio · Tool calling · Thinking · MCP · Coding
Last reviewed: 13 October 2026
When to choose Muse Wonder
Decision guidance for architects—not a feature list.
Best for
- Longer multi-file coding in Muse Code
- Agents that delegate to sub-agents
- Long tasks with parallel tool calls
- Image, video, or audio input with text output
Avoid if
- You want Muse Spark’s lower token price or Contributor-mode discount
- You require open weights or on-prem-only inference
Strengths
Qualitative snapshot for architects—not a public ranking.
- Agentic coding / tools★★★★★
- Multimodal input★★★★★
- API price competitiveness★★☆☆☆
- Open weights★☆☆☆☆
- Ecosystem maturity★★★☆☆
Ecosystem
Built by
- Meta
Muse Wonder is Meta Superintelligence Labs’ newer paid model above Muse Spark.
Competes with
- GPT-5.6
Competes on agentic coding and multimodal tool use via a paid API.
- Claude Opus
Competes for long-running agents and computer-use style workflows.
Works with
- Muse Code
Muse Code can run Muse Wonder for longer tasks. Wonder Full is for subscribers.
Recommended for
- AI Agents
Built for long tasks, sub-agents, and parallel tool calls.
- Llama Models
Clarify open Llama vs closed Muse Wonder and Muse Spark when evaluating Meta stacks.
Often paired with
- Muse Spark
Spark 1.3 stays on the same API at a lower token price, including Contributor mode.
- Llama 4
Llama stays Meta’s open-weight generalist path; Muse Wonder is the closed peak API.
- Muse Glimmer
Glimmer is the Apache-2.0 on-device Muse sibling; Wonder is the paid 1M-context API peak.
How Muse Wonder evolved
Key moments in chronological order.
- Model
Meta releases Muse Wonder (muse-wonder-1.0) on the Meta Model API and in Muse Code, above Muse Spark 1.3. Standard price is $4 / $0.20 cached / $20 per 1M tokens. No Contributor mode on the Model API. 1,048,576-token context; text, image, video, and audio in; text out. Preview. Closed weights. Muse Wonder Full is for Muse Code subscribers and has no separate public token price.
Overview
Capabilities
- Vision: Yes
- Audio: Yes
- Tool calling: Yes
- Thinking: Yes
- MCP: Yes
- Coding: Yes
- Structured output: Yes
Technical specifications
- Provider
- Meta
- License
- Proprietary
- Context window
- 1.0M
- Parameters
- Undisclosed
- Architecture
- Undisclosed (proprietary)
- Release
- 2026-10
- Modalities
- Text, Image, Video, Audio
- Vision
- Yes
- Audio
- Yes
- Tool calling
- Yes
- Thinking
- Yes
- MCP
- Yes
- Open weights
- No
- API
- Yes
- Pricing (input)
- $4 per 1M tokens (muse-wonder-1.0, Standard)
- Pricing (output)
- $20 per 1M tokens
Supported modalities
Text · Image · Video · Audio
Context window
1.0M (1,048,576 tokens)
Pricing
Cached input $0.20 per 1M tokens. Contributor mode is not available for Muse Wonder on the Model API. No long-context premium. Web search grounding is $2.50 per 1,000 queries on top of token cost. Muse Wonder Full has no separate public token price. Verify dev.meta.ai pricing.
Availability
Preview. API id muse-wonder-1.0 at https://api.meta.ai/v1 (OpenAI-compatible chat completions and Anthropic-compatible messages). 1,048,576-token context. Input: text, image, video, and audio. Output: text. Rate limit 3,000 requests/min and 4,000,000 tokens/min per team. Available in Muse Code; Muse Wonder Full is included on Muse Code subscription plans. Consumer Muse app rollout is still upcoming, first to subscribers. Spark examples run on Wonder by changing the model id. Not open weights.
Use cases
- Longer multi-file software engineering in Muse Code
- Agents that plan and delegate to sub-agents
- Long tasks with parallel tool calls
- Multimodal inputs (image, video, audio) with text output
Strengths
- Newer peak above Muse Spark 1.3 on the same Meta Model API
- 1,048,576-token context with no long-context price premium
- Same request shape as Muse Spark: documented Spark examples run by changing the model id to muse-wonder-1.0
Limitations
- Closed weights; preview
- List price is higher than Muse Spark 1.3 ($4/$20 vs $1.25/$4.25)
- No Contributor-mode discount on the Model API
- Muse Wonder Full’s API id and token price are not on the public pricing page
Related guides
Related benchmarks
Related research
- Muse Wonder
Original Paper
- Muse Wonder model card
Original Paper
- Meta Model API pricing
Evaluation
Related tools
Related rankings
Related comparisons
Companies
Explore more models
- Muse SparkMetaMeta Superintelligence Labs’ Muse Spark 1.3 — closed multimodal model for agentic tasks, long-horizon coding, computer use, and 1M-context workflows via Muse Code and the Meta Model API. As of 2026-10-12 it is the lower-priced API under Muse Wonder (muse-wonder-1.0), not Meta’s peak. Succeeds Spark 1.2; max reasoning is still withheld for safety testing.
- Muse GlimmerMetaMeta Superintelligence Labs’ Muse Glimmer — Apache-2.0 ~30B dense multimodal agent model for on-device and single-GPU local agents. Sibling to closed Muse Spark and Muse Wonder; distinct from Llama 4.
- Llama 4MetaMeta’s Llama 4 family — open-weight multimodal models designed for research and commercial use under Meta’s community license.
- Claude FableAnthropicAnthropic’s Claude Fable 5.1 — the most capable widely released Claude for long-horizon agents, deep reasoning, and demanding coding workflows. Mythos 5.1 is the same underlying model with relaxed cyber/life-sciences safeguards for trusted-access programs.
- Claude OpusAnthropicAnthropic’s Claude Opus 5.5 (API id claude-opus-5-5, released 2026-09-22) for complex agentic coding, enterprise work, long-context analysis, and careful instruction following. Anthropic says it matches Fable 5.1 on most work at a lower price. Fable 5.1 remains a separate higher list-price SKU. Sonnet 5.5 (claude-sonnet-5-5, 2026-09-28) is the current Sonnet at the same $2/$10 list price as Sonnet 5. Haiku 5.5 (claude-haiku-5-5, 2026-10-07) is the current Haiku.
- Claude SonnetAnthropicAnthropic’s Claude Sonnet 5.5 (API id claude-sonnet-5-5, released 2026-09-28) for well-scoped everyday coding, agents, and documents. List price matches Sonnet 5 at $2/$10 per 1M tokens. Opus 5.5 remains the high-stakes tier; Haiku 5.5 (2026-10-07) is the low-cost subagent and high-volume tier.