readable AI benchmarks

Arcee AI

Trinity Large Thinking

reasoning · open weights · Apr 1, 2026

Quality Score19.4
Value Score19.9
Reliability28.0
Cache Discount34%

Model specification

Reasoning
reasoning
Input modalities
text
Output modalities
text
Context window
512k tokens
Weights
open weights

Token prices USD per 1M tokens

Input
$0.235
Output
$0.875
Cached Input
$0.155
Cached Output
not available
Cost per task
$0.164

Capability

Intelligence
18.7
Coding
25.8
Agentic
3.7
Omniscience
-44.1
Correct
22.5%
Blended price
$0.243

Answer outcomes

Correct 22.5%Incorrect 66.6%Abstained 10.9%

Artificial Analysis benchmarks

GDPval-AA v2
3%
τ³-Banking
6%
Terminal-Bench v2.1
21%
SciCode
36%
Humanity’s Last Exam
16%
GPQA Diamond
75%
CritPt
1%
AA-Omniscience
28%
AA-LCR
38%

Similar models