readable AI benchmarks (simplified)

readable AI benchmarks

DeepSeek

DeepSeek V3.1 (Reasoning)

reasoning · open weights · Aug 21, 2025

DeepSeek V3.1 (Reasoning) has a Quality Score of 28.0, ranking 272nd among 551 scored models. Per 1M tokens, pricing is $0.590 input, $1.69 output, and $0.590 cached input. It has a 128k token context window. Its Reliability Score is 35.2, ranking 256th among 553 scored models, below average (total average: 35.6). Its Value Score is 20.4, low compared with the total Value average of 30.7.

Quality Score28.0coverage 100.0%: AA Intelligence and AA-Omniscience outcomes available
Value Score20.4
Factual reliability35.2
Cache Discount0%

Model specification

Reasoning
reasoning
Input modalities
none
Output modalities
none
Context window
128k tokens
Weights
open weights

Token prices USD per 1M tokens

Input
$0.590
Output
$1.69
Cached Input
$0.590
Cache write
not available
Cost per task
not available

Capability

Intelligence
13.5
Coding
not available
Agentic
not available
Omniscience
-29.6
Correct
29.0%
Blended price
$0.700

Answer outcomes

Fully graded outcomes (Correct + Incorrect): 87.6%

Correct 29.0%Incorrect 58.6%Partial / not attempted 12.4%

Artificial Analysis benchmarks

Humanity’s Last Exam
14%
GPQA Diamond
78%
CritPt
2%
AA-Omniscience
35%
AA-LCR
57%

Similar models

Support me! Patreon