readable AI benchmarks (simplified)

readable AI benchmarks

OpenAI

GPT-5.4 (xhigh)

reasoning · closed weights · Mar 5, 2026

GPT-5.4 (xhigh) has a Quality Score of 56.4, ranking 62nd among 551 scored models. Per 1M tokens, pricing is $2.50 input, $15.00 output, and $0.250 cached input. It accepts text, image, and pdf input, outputs text, and has a 1050k token context window. Its Reliability Score is 52.9, ranking 84th among 553 scored models, above average (total average: 35.6). Its Value Score is 36.0, high compared with the total Value average of 30.7.

Quality Score56.4coverage 100.0%: AA Intelligence and AA-Omniscience outcomes available
Value Score36.0
Factual reliability52.9
Cache Discount90%

Model specification

Reasoning
reasoning
Input modalities
text, image, pdf
Output modalities
text
Context window
1050k tokens
Weights
closed weights

Token prices USD per 1M tokens

Input
$2.50
Output
$15.00
Cached Input
$0.250
Cache write
not available
Cost per task
not available

Capability

Intelligence
39.0
Coding
not available
Agentic
not available
Omniscience
5.8
Correct
50.8%
Blended price
$2.18

Answer outcomes

Fully graded outcomes (Correct + Incorrect): 95.9%

Correct 50.8%Incorrect 45.1%Partial / not attempted 4.1%

Artificial Analysis benchmarks

GDPval-AA v2
37%
τ³-Banking
40%
Humanity’s Last Exam
44%
GPQA Diamond
92%
CritPt
23%
AA-Omniscience
53%
AA-LCR
82%

Similar models

Support me! Patreon