Meta Muse Spark 1.1

Meta's first proprietary flagship model, succeeding the open-weight Llama line.

Intelligence

AA Intelligence Index51

independent benchmark score from Artificial Analysis, not Timace's own judgment

Best for

  • Multimodal reasoning and tool-use with strong compute efficiency

Specifications

Developer
Meta Superintelligence Labs

confidence 100%

ai.meta.com

Verified Jul 21, 2026 · curated

Context window (tokens)
1M tokens

confidence 85%

ai.meta.com

Verified Jul 21, 2026 · curated

Input $/MTok
$1.25 / MTok

confidence 85%

ai.meta.com

Verified Jul 21, 2026 · curated

Output $/MTok
$4.25 / MTok

confidence 85%

ai.meta.com

Verified Jul 21, 2026 · curated

Modalities
text, image, audio, video, code

confidence 80%

ai.meta.com

Verified Jul 21, 2026 · curated

API available
Yes

confidence 90%

ai.meta.com

Verified Jul 21, 2026 · curated

Open source
No

confidence 90%

ai.meta.com

Verified Jul 21, 2026 · curated

Compare against

Side-by-side →