readable AI benchmarks (simplified)

readable AI benchmarks

Mistral

Ministral 3 8B

non-reasoning · open weights · Dec 2, 2025

Ministral 3 8B has a Quality Score of 12.0, ranking 532nd among 572 scored models. Per 1M tokens, pricing is $0.150 input, $0.150 output, and $0.015 cached input. It accepts text and image input, outputs text, and has a 256k token context window. Its Reliability Score is 15.5, ranking 528th among 574 scored models, below average (total average: 36.5). Its Value Score is 11.2, low compared with the total Value average of 31.5.

Quality Score12.0coverage 100.0%: AA Intelligence and AA-Omniscience outcomes available
Value Score11.2
Factual reliability15.5
Cache Discount90%

Model specification

Reasoning
non-reasoning
Input modalities
text, image
Output modalities
text
Context window
256k tokens
Weights
open weights

Token prices USD per 1M tokens

Input
$0.150
Output
$0.150
Cached Input
$0.015
Cache write
not available
Cost per task
$0.011

Capability

Intelligence
5.5
Coding
not available
Agentic
not available
Omniscience
-69.0
Correct
13.0%
Blended price
$0.055

Answer outcomes

Fully graded outcomes (Correct + Incorrect): 94.9%

Correct 13.0%Incorrect 81.9%Partial / not attempted 5.1%

Artificial Analysis benchmarks

GDPval-AA v2
0%
τ³-Banking
4%
SciCode
21%
Humanity’s Last Exam
4%
GPQA Diamond
47%
CritPt
0%
AA-Omniscience
16%
AA-LCR
26%

Similar models

Support me! Patreon