readable AI benchmarks (simplified)

readable AI benchmarks

Meta

Muse Spark 1.3 (xhigh)

reasoning · closed weights · Sep 2, 2026

Muse Spark 1.3 (xhigh) has a Quality Score of 69.6, ranking 20th among 277 scored models. Per 1M tokens, pricing is $1.25 input, $4.25 output, and $0.150 cached input. It accepts text, image, and video input, outputs text, and has a 1M token context window. Its Reliability Score is 61.6, ranking 24th among 277 scored models, above average (total average: 34.6). Its Value Score is 51.7, high compared with the total Value average of 28.7.

Quality Score69.6coverage 91.2%: missing Epoch General ECI, CursorBench 3.2 and DeepSWE v1.1
Value Score51.7
Reliability61.6
Cache Discount88%

Model specification

Reasoning
reasoning
Input modalities
text, image, video
Output modalities
text
Context window
1M tokens
Weights
closed weights

Token prices USD per 1M tokens

Input
$1.25
Output
$4.25
Cached Input
$0.150
Cached Output
not available
Cost per task
$0.547

Capability

Intelligence
60.8
Coding
76.5
Agentic
56.1
Omniscience
23.1
Correct
41.5%
Blended price
$0.780

Answer outcomes

Attempt rate: 59.9%

Correct 41.5%Incorrect 18.4%Abstained 40.1%

Artificial Analysis benchmarks

GDPval-AA v2
60%
τ³-Banking
47%
Terminal-Bench v2.1
85%
SciCode
59%
Humanity’s Last Exam
47%
GPQA Diamond
94%
CritPt
26%
AA-Omniscience
62%
AA-LCR
79%

Similar models