readable AI benchmarks

Upstage

Solar Pro 4

reasoning · closed weights · Aug 6, 2026

Quality Score43.9
Value Score37.9
Reliability49.6
Cache Discount80%

Model specification

Reasoning
reasoning
Input modalities
text
Output modalities
text
Context window
512k tokens
Weights
closed weights

Token prices USD per 1M tokens

Input
$0.300
Output
$1.20
Cached Input
$0.060
Cached Output
not available
Cost per task
$0.223

Capability

Intelligence
41.6
Coding
52.7
Agentic
33.6
Omniscience
-0.8
Correct
18.9%
Blended price
$0.222

Answer outcomes

Correct 18.9%Incorrect 19.8%Abstained 61.3%

Artificial Analysis benchmarks

GDPval-AA v2
39%
τ³-Banking
23%
Terminal-Bench v2.1
57%
SciCode
44%
Humanity’s Last Exam
29%
GPQA Diamond
89%
CritPt
5%
AA-Omniscience
50%
AA-LCR
71%

Similar models