readable AI benchmarks

Tencent

Hy3

reasoning · open weights · Jul 6, 2026

Quality Score44.0
Value Score44.2
Reliability40.8
Cache Discount75%

Model specification

Reasoning
reasoning
Input modalities
text
Output modalities
text
Context window
256k tokens
Weights
open weights

Token prices USD per 1M tokens

Input
$0.136
Output
$0.554
Cached Input
$0.034
Cached Output
not available
Cost per task
$0.036

Capability

Intelligence
42.2
Coding
58.8
Agentic
31.4
Omniscience
-18.5
Correct
32.0%
Blended price
$0.106

Answer outcomes

Correct 32.0%Incorrect 50.4%Abstained 17.6%

Artificial Analysis benchmarks

GDPval-AA v2
36%
τ³-Banking
23%
Terminal-Bench v2.1
64%
SciCode
48%
Humanity’s Last Exam
33%
GPQA Diamond
90%
CritPt
5%
AA-Omniscience
41%
AA-LCR
75%

Similar models