ModelsCompareBest forBenchmarksStatusPricingAPI

Hy3

Tencent Hunyuan·Pending·Released 2026-07-06text
Context256k
Max output-
Input / 1M-
Output / 1M-
Cached in-
Benchmarks41

Benchmarks

Cited public sources · rank is exact-benchmark, apples-to-apples

LMArena Elo#22/621440.6 ↗
AA-LCR#2/473.4% ↗
Apex-Agent25.6% ↗
ArxivMath52.2% ↗
BrowseComp#5/1584.2% ↗
CL-bench23.8% ↗
CL-bench Life17% ↗
ClawEval68.5% ↗
CMT-Benchmark37.8% ↗
DeepSearchQA91% ↗
DeepSWE28% ↗
e-bench50.2% ↗
FrontierScience-Olympiad74.8% ↗
FrontierScience-Research21.3% ↗
GPQA Diamond#20/7090.4% ↗
HLE (with tools, text-only)#1/253.2% ↗
HorizonMath7.1% ↗
Humanity's Last Exam (no tools)#10/2937% ↗
Hy-Backend 2.025% ↗
Hy-CompanyBench41.7% ↗
Hy-Euler Pro24.2% ↗
Hy-FinModelBench69% ↗
Hy-Math60.9% ↗
Hy-SkillsWorld45.8% ↗
Hy-SWE Max49% ↗
IMOAnswerBench#1/290% ↗
MathArena Apex38.7% ↗
MCP Atlas#2/979.1% ↗
NL2Repo#4/445.6% ↗
PHYBench77.4% ↗
ProdBench23% ↗
SkillsBench55.3% ↗
SuperChem54.9% ↗
SWE-bench Multilingual#5/775.8% ↗
SWE-bench Pro#3/757.9% ↗
SWE-bench Verified#7/4578% ↗
Terminal-Bench 2.1#10/1771.7% ↗
Toolathlon#4/648.5% ↗
USAMO 202672% ↗
WideSearch76.4% ↗
WildClawBench53.6% ↗

Compare

vs GPT-6 Astravs GPT-5.6 Solvs GPT-5.6 Terravs GPT-5.6 Lunavs Claude Fable 5.1vs Claude Opus 5vs Claude Sonnet 5vs Claude Haiku 4.5

Access Hy3 and every other model through one endpoint with automatic failover: Respan gateway.