Skip to main content
AI Choice Engine

Meta

Muse Wonder 1.0

Muse Wonder 1.0 is a Meta model with 1.05M context, $4 in / $20 out per million tokens, last verified October 25, 2026. Open weights: not verified. Public-board scores are linked to the publisher; vendor-only figures stay in the watch-out, not on the board.

Meta’s hosted preview for complex software engineering and long-running multimodal agents, separate from the subscription-only Wonder Full variant.

Recorded profile verification October 25, 2026Individual provider fields may changeAccess: preview — The publisher model card dates the Wonder family release October 12,2026. Current docs label API Wonder preview, with 1,048,576-token context. Country restrictions apply; Wonder is not currently available in South Korea. No downloadable Wonder weights/licence or exact parameter count is established.

AI model specification details
SpecificationMuse Wonder 1.0
ProviderMeta
TierFrontier
Context window1.05M
Max outputNot verifiedUnverified
Input / 1M tokens$4
Output / 1M tokens$20
WeightsNot verifiedUnverified
ParametersNot verifiedUnverified
Reasoning levelsminimal, low, medium, high, xhigh, max
Modalitiestext, image, video, audio, file
API model idmuse-wonder-1.0
ReleasedOctober 12, 2026

Pricing tiers: Meta Model API Standard USD/MTok input/output $4/$20, cached input $0.20. No long-context premium; web search adds $2.50/1,000 queries. API Contributor mode is unavailable for Wonder. Team limit3,000 RPM/4M TPM; code-subscription quotas are a separate contract. Muse Wonder Full has no Model API token tariff and is not selected by setting this API model to max.

Best for

  • Evaluating long-running agent workflows
  • Complex multi-file coding
  • Multimodal tool-using assistants

Watch out

Reasoning minimal/low/medium/high/xhigh/max; none returns400 and omitted effort is model-determined. Reasoning consumes output budget and billed output tokens. Standard mode content is not used for training; Code subscriptions have a separate Contributor-mode default and data-use contract. Wonder Full is a different higher-capability model, not max effort on this API. A hard output cap, exact parameters and native-weight availability remain unknown. Publisher benchmark tables are not imported as independent observations.

History

Muse Wonder 1.0 timeline

Release and status events for this model that passed two-source verification — dated, sourced, and linked.

Source receipts

Catalog figures for this model were checked against the following sources.

Common questions

Muse Wonder 1.0

Answered from the verified figures on this page rather than general guidance.

How much does Muse Wonder 1.0 cost per million tokens?
Muse Wonder 1.0 lists base rates of $4 per million input tokens and $20 per million output tokens. Estimate both input and output token volume and any additional charges for your workload. Meta Model API Standard USD/MTok input/output $4/$20, cached input $0.20. No long-context premium; web search adds $2.50/1,000 queries. API Contributor mode is unavailable for Wonder. Team limit3,000 RPM/4M TPM; code-subscription quotas are a separate contract. Muse Wonder Full has no Model API token tariff and is not selected by setting this API model to max.
What is Muse Wonder 1.0's context window?
Muse Wonder 1.0 accepts about 1.05M tokens of context. That only matters if you routinely send very long documents, large codebases, or multi-turn histories that approach that limit.
What is Muse Wonder 1.0 best for?
Muse Wonder 1.0 is a frontier tier from Meta. It suits evaluating long-running agent workflows, complex multi-file coding, multimodal tool-using assistants. Reasoning minimal/low/medium/high/xhigh/max; none returns400 and omitted effort is model-determined. Reasoning consumes output budget and billed output tokens. Standard mode content is not used for training; Code subscriptions have a separate Contributor-mode default and data-use contract. Wonder Full is a different higher-capability model, not max effort on this API. A hard output cap, exact parameters and native-weight availability remain unknown. Publisher benchmark tables are not imported as independent observations.
Model comparison catalogBenchmark library