Muse Spark 1.2 Review
Meta's model that launched alongside the Muse Code terminal agent, with a low price and a score that still sits in the top group.
Scores
Score by area
Summary
Strengths and weaknesses
Strengths
- AA Index 57 at only $1.25/$4.25
- Comes with Muse Code, an agent that runs in the terminal
- Scores 80 to 82.9% on Terminal-Bench 2.1, 1.05M context
Weaknesses
- Only released on August 5, 2026, with little real-world validation so far
- Muse Code is still immature compared with OpenAI's and Anthropic's tools
- Falls behind the top group on the hardest reasoning tasks
Review
Detailed review
Ranking source
Updated August 9, 2026 from the Artificial Analysis Intelligence Index (v4.1 scale), cross-checked against each provider's release notes and official pages. Scores on the v4.1 scale can't be compared directly with earlier rankings.
When to use it
Use it when you need a low-cost coding agent: fixing repetitive bugs, writing scripts, and running supervised terminal tasks.
Notes
AA Intelligence Index: 57 (xhigh); context 1.05M; released August 5, 2026; price $1.25/$4.25. Its price-to-score ratio is among the best in this table.
Ông Chú MMO
AI Reviewer
Reviews are written by the Ông Chú MMO team based on hands-on experience.
Comments
No comments yet.
Similar tools
Compare toolsMultimodal LLM
Qwen3.8-Max
Truly multimodal: text, images, and video in a single model, at a very friendly price.
9.2Multimodal LLM
Gemini 3.1 Pro
Google's highest Pro version, very strong at vision but now lagging in overall score.
8.3Balanced LLM
GPT-5.6 Terra
The mid-tier version of GPT-5.6, scoring close to Sol for just over a third of the price.
9.1
