Superbash

Muse Spark 1.2

A provisional B-tier local/open option that is not winning on intelligence or price.

Canonical model record

Current identity, limits, and pricing

Provider source · checked 2026-08-14 ↗

Status: Current
API model ID: Not publicly verified
Context: 1M
Max output: Not publicly verified
API price / 1M tokens: Not publicly verified

Visual prompt runs

Benchmark runs

Open each generated scene, or compare the same prompt across models.

0 runs

Superbash commentary

Our take

Back to the full tier list →

Muse Spark 1.2 remains a provisional B-tier choice. Meta is trying to regain ground with an open local option, but Muse is not yet the intelligence leader and DeepSeek is a stronger price story.

Best for

Supervised coding workflows
Specialist use with a stronger reviewer

Watch out

Limited firsthand team testing
Not currently the intelligence leader
DeepSeek is the stronger budget comparison
Treat the B placement as provisional

Why it is ranked here

Muse remains in B provisionally because the team has not completed enough firsthand testing to justify a stronger move.
The team sees local availability as its clearest strategic angle rather than a decisive intelligence lead.
This is a provisional routing recommendation while detailed firsthand testing is still being documented.

Evidence and commentary

2026-08-17

Superbash editorial model ranking

Takeaway: Muse Spark 1.2 is currently placed in Tier B.

The August 2026 editorial roster places Muse Spark 1.2 at rank 15.

Open source →

2026-08-05

Meet Muse Spark 1.2 and Muse Code

Takeaway: Meta’s current developer material verifies Muse Spark 1.2 as the coding model behind Muse Code.

The source confirms the current model identity; fields not exposed in crawlable public documentation remain explicitly blank in the catalog.