readable AI benchmarks (simplified)

readable AI benchmarks

Multiverse Computing

Quasar 438B

reasoning · closed weights · Aug 10, 2026

Quasar 438B has a Quality Score of 43.7, ranking 67th among 267 scored models. Per 1M tokens, pricing is $0.600 input and $1.80 output. It accepts text input, outputs text, and has a 1M token context window. Its Reliability Score is 48.7, ranking 53rd among 267 scored models, above average (total average: 33.5). Its Value Score is 33.1, high compared with the total Value average of 27.7.

Quality Score43.7coverage 91.2%: missing Epoch General ECI, CursorBench 3.2 and DeepSWE v1.1
Value Score33.1
Reliability48.7
Cache Discountnot available

Model specification

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

Token prices USD per 1M tokens

Input
$0.600
Output
$1.80
Cached Input
not available
Cached Output
not available
Cost per task
$0.596

Capability

Intelligence
43.0
Coding
61.2
Agentic
36.9
Omniscience
-2.6
Correct
15.5%
Blended price
$0.720

Answer outcomes

Attempt rate: 33.6%

Correct 15.5%Incorrect 18.1%Abstained 66.4%

Artificial Analysis benchmarks

GDPval-AA v2
41%
τ³-Banking
29%
Terminal-Bench v2.1
69%
SciCode
45%
Humanity’s Last Exam
19%
GPQA Diamond
73%
CritPt
9%
AA-Omniscience
49%
AA-LCR
75%

Similar models