Muse Spark 1.2

meta/muse-spark-1.2

coding-optimized for agentic workflows and large codebases.

cmd --model meta/muse-spark-1.2
Intelligence index
54.1
Output speed
not yet scored
Input
$1.25 /M
Output
$4.25 /M
Cache read
$0.15 /M
Agent-loop cost
$0.48 /M in
Context window
1.05M tokens
Released
August 5, 2026
Modalities

vs. the lineup

Muse Spark 1.2 beside its stablemates and nearest rivals. The ◆ marks the best value in each column across every row shown.

pin a rival:
ModelIntelligenceCodingSpeedInput $/MOutput $/MBlended $/MContext
GPT-5.6 Terra5576.7144.3$1$6$2.251.05M
GPT-5.554.874.9$5$30$11.25200K
Muse Spark 1.254.172.2$1.25$4.25$21.05M
Muse Spark 1.2 Contributor54.172.2$0.10$0.20$0.131.05M
Muse Spark 1.150.671.3166.2$1.25$4.25$21.05M
Step 3.7 Flash30.339.6392.5$0.20$1.15$0.44256K
DeepSeek V4 Propinned44.359.459.6$0.43$0.87$0.541M
DeepSeek V4 Flash (latest)pinned5069.1102.2$0.14$0.28$0.181M

coding performance

The Intelligence Index and its sub-scores, ranked against every scored model in the catalog. A metric that has not been measured for Muse Spark 1.2 has been left empty.

Coding Index
72.2
#10 of 40 scored
Terminal-Bench
80.1
#12 of 40 scored
Intelligence Index
54.1
#8 of 46 scored
Long-context reasoning
64.7
reasoning across a long context
SciCode
56.4
scientific coding
GPQA Diamond
90.4
graduate-level QA

usage calculator

How far a month of credits goes on Muse Spark 1.2.

Input tokensfresh prompt
800
Output tokensmodel reply
180
Cache read tokensre-read context
50K
cost / request $0.0093 · in $1.25 · out $4.25 · cache $0.15 per M
fresh input 11%output 8%cache reads 81%
Requests / 30 days
1.1K
$10 credits ÷ $0.0093 per request
~216 quick fixes~43 bug fixes~7 feature PRs

what real work costs

Real coding tasks priced end to end on Muse Spark 1.2, from a quick lookup to a full-repo agent run.

One agent task
$0.13
180K in at 75% cache hit, 12K out
What you pay
$0.66 /M
all-in across every token that task touched
Sticker input
$1.25 /M
cache reads bill at $0.15 /M instead
TaskTokens in · outMuse Spark 1.2Muse Spark 1.2 ContributorClaude Haiku 4.5
Quick lookup / one-liner8K · 1K$0.0063$0.0003$0.0065
Review a 500-line PR60K · 4K$0.05$0.0027$0.04
Fix a bug (agent loop)180K · 12K$0.13$0.0072$0.12
Refactor a module320K · 20K$0.23$0.01$0.21
Full-repo agent run900K · 45K$0.54$0.03$0.49

frequently asked

What is Muse Spark 1.2 best for?+
Muse Spark 1.2 is a reasoning model from Meta, designed for complex agentic tasks. It accepts text, images, video, audio, and PDF documents, returns text, and offers a 1M-token context window. The model is built to support multi-agent workflows, whether as either a main agent that plans and delegates or as a subagent executing in parallel. It works across multiple coding harnesses and supports structured output, parallel function calling, and configurable reasoning effort. In Meta’s testing, it performs well on multi-file refactors, extended debugging sessions, whole-repository generation, and tasks that stretch well past a single prompt.
How much does Muse Spark 1.2 cost?+
$1.25/M input and $4.25/M output, cache reads $0.15/M. In an agent loop most input is cache-read, so the effective input rate is about $0.48/M.
Which plan do I need?+
Available on GOAT and above.
How do I switch to it?+
Run cmd --model meta/muse-spark-1.2, or type /model in a session and pick it. You can switch mid-session without losing context.

Ship code that matches your taste

Command Code is the AI coding agent that continuously learns your taste. Start for $1.

Benchmarks from Artificial Analysis (v4.1)commandcode.ai/models/muse-spark-1-2