AI BENCHY
Advertise here
#25

Muse Spark 1.3

Meta Release: 2026-09-02 Tested on: 2026-09-02 21:47 meta/muse-spark-1.3::low
~1T total (~50B active)MoEClosedEstimated
(high) (medium) (low)

Summary

Muse Spark 1.3 scores 9.0 on AI BENCHY and ranks #25. It has 9.6 reliability, a 83.3% pass rate, $0.633 total cost, and 43.67s average response time.

Model facts

Researched on 2026-09-02

Estimated
Parameters
~1T total (~50B active)
Architecture
MoE
Availability
Closed
License
-

Best estimate from public evidence; the vendor did not disclose every value. API-only exact checkpoint. Meta does not disclose parameter counts or architecture; estimates follow the Muse Spark 1.2 family record.

Parameter source Availability source

Score

9.0

Consistency

9.0

Reliability

9.6

Total Cost (Current Price)

$0.633

Total Output Tokens

117,915

Total Input Tokens

104,966

Input Price

$1.250 / 1M

Output Price

$4.250 / 1M

Tests Correct

Wrong Tests: 5

Attempt pass rate: 83.3%

Flaky tests

3

Flaky tests had mixed outcomes across runs (at least one pass and one fail).

Response Time (avg)

43.67s

Response Time (max): 190.38s

Response Time (total): 960.70s

Wrong answer: 3 API error: 1 Invalid tool call: 1

Hamster playing table tennis

Prompt: Create a detailed SVG illustration of a hamster playing table tennis.

#25 Muse Spark 1.3

low
Cost
$0.017
Time
83.9s
Tokens
3,959 tok

Charts

Choose the first model, then click a second model to open a side-by-side page.

Total Output Tokens

Score vs Total Output Tokens

Quick Compare

Muse Spark 1.3lowvsMuse Spark 1.3highMuse Spark 1.3lowvsGrok 4.5highMuse Spark 1.3lowvsQwen3.8 MaxmediumMuse Spark 1.3lowvsMuse Spark 1.3mediumMuse Spark 1.3lowvsGPT-5.5mediumMuse Spark 1.3lowvsGPT-5.3-CodexmediumMuse Spark 1.3lowvsGemini 3.5 FlashmediumMuse Spark 1.3lowvsSeed 2.1 TurbohighMuse Spark 1.3lowvsClaude Fable 5.1highMuse Spark 1.3lowvsGemini 3.5 Flashlow

Category Breakdown

Category Score Consistency Tests Correct
Anti-AI Tricks 10.0 10.0
Coding 10.0 10.0
Combined 8.7 6.9
Data parsing and extraction 10.0 10.0
Domain specific 2.9 4.4
General Intelligence 10.0 10.0
Instructions following 9.8 10.0
Puzzle Solving 10.0 10.0
Tool Calling 10.0 10.0
Trivia 3.0 10.0

Compared models