readable AI benchmarks

OpenAI

GPT-5.6 Luna (high)

reasoning · closed weights · Jul 9, 2026

Quality Score52.8
Value Score48.6
Reliability44.0
Cache Discount90%

Model specification

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

Token prices USD per 1M tokens

Input
$0.200
Output
$1.20
Cached Input
$0.020
Cached Output
$0.250
Cost per task
$0.022

Capability

Intelligence
47.0
Coding
63.3
Agentic
41.0
Omniscience
-12.0
Correct
41.8%
Blended price
$0.174

Answer outcomes

Correct 41.8%Incorrect 53.8%Abstained 4.4%

Artificial Analysis benchmarks

GDPval-AA v2
48%
τ³-Banking
25%
Terminal-Bench v2.1
70%
SciCode
51%
Humanity’s Last Exam
33%
GPQA Diamond
89%
CritPt
17%
AA-Omniscience
44%
AA-LCR
74%

Similar models