MetaAgentic MultimodalPreview

Muse Spark 1.2

Meta's current Muse Spark coding model for agentic workflows, computer use, and native multimodal perception.

Facts Verified
Aug 25, 2026
Status
Preview
Profile
Source-backed
Compare ModelOfficial Docs
API
Available

Specifications

Context
Not published
Max output
Not published
CompanyMeta
Release DateAug 5, 2026
StatusPreview
License / AvailabilityDirect API available

Capability Map

Verified fields
Reasoning
Not published
Vision
Supported
Audio
Supported
Video
Supported
API
Supported
Open weights
Not published
Input modalities
textimageaudiovideodocument
Output: text

Developer API Rates

Direct token rate

No public direct API token price

The verified record does not store a fixed direct token rate.

Direct token pricing is intentionally omitted until the current Meta Model API first-party rate card can be verified directly.

Calculate Usage in AI Price Tracker →

Verified Capabilities

  • Multimodal input: text, image, audio, video, document
  • Direct API availability is published

Pricing & Availability Notes

  • Preview status; availability may change
  • Available through Meta Model API public preview and Muse Code.
  • Canonical context size is intentionally omitted until it is directly verified from current first-party developer documentation.
  • Direct token pricing is intentionally omitted until the current Meta Model API first-party rate card can be verified directly.

Sources & Freshness

Verified Aug 25, 2026

Available through Meta Model API public preview and Muse Code.

Canonical context size is intentionally omitted until it is directly verified from current first-party developer documentation.