Back to Models

Meta: Muse Spark 1.1

Metatext
9.1 / 10 Overall Rating

Meta's Muse Spark 1.1 is a 1M-token context multimodal reasoning model engineered for agentic workflows across text, image, audio, and video inputs. At $1.25/1M input and $4.25/1M output tokens, it significantly undercuts GPT-4o and Claude 3.5 Sonnet on cost while providing a 943K max output token limit. It is targeted at heavy context ingestion, video/audio processing, and agent orchestration.

Context Window
1049K
Knowledge Cutoff
2025
Max Output
16K
License
Open Weights

Where it Excels (Pros)

  • 1M token context window
  • 943K max output limit
  • Cheaper than GPT-4o and Sonnet

Limitations (Cons)

  • Text only output generation
  • Unproven coding capabilities vs Sonnet

Benchmark Breakdown

Reasoning (MMLU)82%
Coding (HumanEval)78%
Mathematics (MATH)76%
Long-Context Retrieval95%

Pricing Matrix

Input cost / 1M tokens$1.25
Output cost / 1M tokens$4.25

Compare Meta: Muse Spark 1.1 with other frontier models:

Related Skills, Tools & Automations