readable AI benchmarks (simplified)

readable AI benchmarks

SpaceXAI

Grok 4.20 0309 (Non-reasoning)

non-reasoning · closed weights · Mar 10, 2026

Grok 4.20 0309 (Non-reasoning) has a Quality Score of 25.6, ranking 301st among 551 scored models. Per 1M tokens, pricing is $2.00 input, $6.00 output, and $0.200 cached input. It accepts text, image, and pdf input, outputs text, and has a 2M token context window. Its Reliability Score is 28.7, ranking 323rd among 553 scored models, below average (total average: 35.6). Its Value Score is 17.6, low compared with the total Value average of 30.7.

Quality Score25.6coverage 100.0%: AA Intelligence and AA-Omniscience outcomes available
Value Score17.6
Factual reliability28.7
Cache Discount90%

Model specification

Reasoning
non-reasoning
Input modalities
text, image, pdf
Output modalities
text
Context window
2M tokens
Weights
closed weights

Token prices USD per 1M tokens

Input
$2.00
Output
$6.00
Cached Input
$0.200
Cache write
not available
Cost per task
not available

Capability

Intelligence
14.6
Coding
not available
Agentic
not available
Omniscience
-42.6
Correct
26.1%
Blended price
$1.14

Answer outcomes

Fully graded outcomes (Correct + Incorrect): 94.7%

Correct 26.1%Incorrect 68.6%Partial / not attempted 5.3%

Artificial Analysis benchmarks

Humanity’s Last Exam
25%
GPQA Diamond
78%
CritPt
0%
AA-Omniscience
29%
AA-LCR
25%

Similar models

Support me! Patreon