Tencent model product
Hy3 is a 295B-parameter Mixture-of-Experts (MoE) model from Tencent with 21B active parameters and a 256K context window.
Updated Sep 8, 2026. Default version: Hy3
This profile uses the model's current scored version. Arena ratings and prices are shown separately.
Benchmark scores for Hy3.
Benchmark | Score | Rank | Participants | Percentile | Evidence | Evaluated |
|---|
| BenchmarkCL-bench | Score23.80% | Rank01 | Participants2 | Percentile100.00% | EvidenceC | Evaluated |
| BenchmarkCL-bench (Life) | Score17.00% | Rank01 | Participants1 | Percentile100.00% | EvidenceC | Evaluated |
| BenchmarkCMT-Benchmark | Score37.90% | Rank01 | Participants1 | Percentile100.00% | EvidenceC | Evaluated |
| BenchmarkHumanity's Last Exam (no tools, text-only) | Score47.00% | Rank01 | Participants3 | Percentile100.00% | EvidenceC | Evaluated |
| BenchmarkPHYBench | Score77.40% | Rank01 | Participants1 | Percentile100.00% | EvidenceC | Evaluated |
| BenchmarkAA-LCR | Score73.40% | Rank02 | Participants19 | Percentile94.44% | EvidenceC | Evaluated |
| BenchmarkHorizonMath | Score7.10% | Rank02 | Participants4 | Percentile66.67% | EvidenceC | Evaluated |
| BenchmarkArXivMath | Score52.20% | Rank03 | Participants3 | Percentile0.00% | EvidenceC | Evaluated |
| BenchmarkFrontierScience Olympiad | Score74.80% | Rank03 | Participants3 | Percentile0.00% | EvidenceC | Evaluated |
| BenchmarkIMO-AnswerBench | Score90.00% | Rank03 | Participants20 | Percentile89.47% | EvidenceC | Evaluated |
| BenchmarkUSAMO 2026 | Score30.24 points | Rank03 | Participants3 | Percentile0.00% | EvidenceC | Evaluated |
| BenchmarkWildClawBench | Score53.60% | Rank03 | Participants5 | Percentile50.00% | EvidenceC | Evaluated |
| BenchmarkClaw-Eval | Score68.50% | Rank04 | Participants14 | Percentile76.92% | EvidenceC | Evaluated |
| BenchmarkDeepSearchQA | Score91.00% | Rank04 | Participants10 | Percentile66.67% | EvidenceC | Evaluated |
| BenchmarkFrontierScience Research | Score21.30% | Rank04 | Participants4 | Percentile0.00% | EvidenceC | Evaluated |
| BenchmarkSkillsBench | Score55.30% | Rank04 | Participants9 | Percentile62.50% | EvidenceC | Evaluated |
| BenchmarkSuperChem | Score54.90% | Rank04 | Participants4 | Percentile0.00% | EvidenceC | Evaluated |
| BenchmarkHumanity's Last Exam (with tools, text-only) | Score53.20% | Rank05 | Participants5 | Percentile0.00% | EvidenceC | Evaluated |
| BenchmarkWideSearch | Score76.40% | Rank05 | Participants11 | Percentile60.00% | EvidenceC | Evaluated |
| BenchmarkMathArena Apex | Score38.70% | Rank06 | Participants9 | Percentile37.50% | EvidenceC | Evaluated |
| BenchmarkAPEX-Agents | Score25.60% | Rank09 | Participants9 | Percentile0.00% | EvidenceC | Evaluated |
| BenchmarkMCP Atlas | Score79.10% | Rank09 | Participants34 | Percentile75.76% | EvidenceC | Evaluated |
| BenchmarkDeepSWE | Score28.00% | Rank12 | Participants13 | Percentile8.33% | EvidenceC | Evaluated |
| BenchmarkNL2Repo | Score45.60% | Rank13 | Participants22 | Percentile42.86% | EvidenceC | Evaluated |
| BenchmarkSWE-bench Multilingual | Score75.80% | Rank13 | Participants43 | Percentile71.43% | EvidenceC | Evaluated |
| BenchmarkBrowseComp | Score84.20% | Rank15 | Participants63 | Percentile77.42% | EvidenceC | Evaluated |
| BenchmarkSWE-Bench Verified | Score78.00% | Rank22 | Participants113 | Percentile81.25% | EvidenceC | Evaluated |
| BenchmarkTerminal-Bench 2.1 | Score71.70% | Rank22 | Participants35 | Percentile38.24% | EvidenceC | Evaluated |
| BenchmarkSWE-Bench Pro | Score57.90% | Rank24 | Participants55 | Percentile57.41% | EvidenceC | Evaluated |
| BenchmarkGPQA | Score90.40% | Rank25 | Participants247 | Percentile90.24% | EvidenceC | Evaluated |
Preference and agent-evaluation results for the default version.
Arena | Category | Rank | Rating / score | Votes | Observations | Result date |
|---|
| Arenatext | Categoryindustry mathematical | Rank13 | Rating / score1,493.94 | Votes387 | ObservationsN/A | Result date |
| Arenatext style control | Categoryindustry mathematical | Rank13 | Rating / score1,504.54 | Votes387 | ObservationsN/A | Result date |
| Arenatext factuality | Categoryindustry life and physical and social science | Rank19 | Rating / score1,491.86 | Votes1,035 | ObservationsN/A | Result date |
| Arenatext style control | Categoryindustry life and physical and social science | Rank19 | Rating / score1,493.88 | Votes1,050 | ObservationsN/A | Result date |
| Arenatext factuality | Categoryrussian | Rank23 | Rating / score1,467.54 | Votes709 | ObservationsN/A | Result date |
| Arenatext style control | Categoryindustry medicine and healthcare | Rank23 | Rating / score1,488.16 | Votes459 | ObservationsN/A | Result date |
| Arenaagent steerability | Categoryoverall | Rank25 | Rating / score0.00 | VotesN/A | Observations11.8K | Result date |
| Arenatext | Categorymath | Rank26 | Rating / score1,468.12 | Votes359 | ObservationsN/A | Result date |
| Arenatext style control | Categorychinese | Rank26 | Rating / score1,516.58 | Votes465 | ObservationsN/A | Result date |
| Arenatext style control | Categorymath | Rank28 | Rating / score1,475.24 | Votes359 | ObservationsN/A | Result date |
| Arenawebdev | Categorywebdev-react | Rank29 | Rating / score1,517.75 | Votes2,224 | ObservationsN/A | Result date |
| Arenaagent bash recovery steps | Categoryoverall | Rank32 | Rating / score-0.01 | VotesN/A | Observations22.6K | Result date |
| Arenatext style control | Categoryspanish | Rank32 | Rating / score1,461.26 | Votes268 | ObservationsN/A | Result date |
| Arenaagent praise complaint | Categoryoverall | Rank33 | Rating / score-0.04 | VotesN/A | Observations3.7K | Result date |
| Arenatext | Categoryindustry life and physical and social science | Rank33 | Rating / score1,466.80 | Votes1,050 | ObservationsN/A | Result date |
| Arenatext style control | Categoryfrench | Rank33 | Rating / score1,484.93 | Votes302 | ObservationsN/A | Result date |
| Arenawebdev | Categoryoverall | Rank33 | Rating / score1,511.65 | Votes3,173 | ObservationsN/A | Result date |
| Arenawebdev | Categorywebdev | Rank33 | Rating / score1,511.65 | Votes3,173 | ObservationsN/A | Result date |
| Arenaagent | Categoryoverall | Rank34 | Rating / score-0.02 | VotesN/A | Observations876.8K | Result date |
| Arenatext | Categorychinese | Rank35 | Rating / score1,499.75 | Votes465 | ObservationsN/A | Result date |
| Arenatext | Categoryrussian | Rank36 | Rating / score1,451.41 | Votes745 | ObservationsN/A | Result date |
| Arenatext style control | Categoryexpert | Rank36 | Rating / score1,489.67 | Votes709 | ObservationsN/A | Result date |
| Arenatext style control | Categoryrussian | Rank36 | Rating / score1,465.70 | Votes745 | ObservationsN/A | Result date |
| Arenatext | Categoryindustry medicine and healthcare | Rank37 | Rating / score1,459.36 | Votes459 | ObservationsN/A | Result date |
| Arenatext | Categoryfrench | Rank39 | Rating / score1,463.39 | Votes302 | ObservationsN/A | Result date |
| Arenatext | Categoryspanish | Rank39 | Rating / score1,446.96 | Votes268 | ObservationsN/A | Result date |
| Arenaagent task outcome explicit | Categoryoverall | Rank40 | Rating / score-0.06 | VotesN/A | Observations11.7K | Result date |
| Arenawebdev | Categorywebdev-html | Rank41 | Rating / score1,476.61 | Votes350 | ObservationsN/A | Result date |
| Arenatext style control | Categoryenglish | Rank42 | Rating / score1,467.69 | Votes2,697 | ObservationsN/A | Result date |
| Arenatext style control | Categorylonger query | Rank43 | Rating / score1,466.00 | Votes3,165 | ObservationsN/A | Result date |
Official vendor API pricing appears first, followed by individual provider offers.
Provider | Provider model ID | Region | Input / 1M | Output / 1M | Context | Updated |
|---|
| ProviderTencent TokenHub | Provider model IDhy3 | Regionglobal | Input / 1MN/A | Output / 1MN/A | Context256K | Updated |
| ProviderKenari | Provider model IDhy3 | Regionglobal | Input / 1MN/A | Output / 1MN/A | Context256K | Updated |
| ProviderTencent Token Plan | Provider model IDhy3 | Regionglobal | Input / 1MN/A | Output / 1MN/A | Context256K | Updated |
| ProviderNanoGPT | Provider model IDtencent/hy3 | Regionglobal | Input / 1M$0.066 | Output / 1M$0.26 | Context262.1K | Updated |
| ProviderOpenRouter | Provider model IDtencent/hy3 | Regionglobal | Input / 1M$0.132 | Output / 1M$0.528 | Context262.1K | Updated |
| ProviderKilo Gateway | Provider model IDtencent/hy3 | Regionglobal | Input / 1M$0.132 | Output / 1M$0.528 | Context262.1K | Updated |
| ProviderJalapeno Cloud | Provider model IDHy3 | Regionglobal | Input / 1M$0.14 | Output / 1M$0.58 | Context202.8K | Updated |
| ProviderDeep Infra | Provider model IDtencent/Hy3 | Regionglobal | Input / 1M$0.14 | Output / 1M$0.58 | Context262.1K | Updated |
| ProviderHugging Face | Provider model IDtencent/Hy3 | Regionglobal | Input / 1M$0.14 | Output / 1M$0.58 | Context262.1K | Updated |
| ProviderOpenCode Go | Provider model IDhy3 | Regionglobal | Input / 1M$0.14 | Output / 1M$0.58 | Context256K | Updated |
| ProviderRequesty | Provider model IDhy3 | Regionglobal | Input / 1M$0.14 | Output / 1M$0.58 | Context262.1K | Updated |
| ProviderVercel AI Gateway | Provider model IDtencent/hy3 | Regionglobal | Input / 1M$0.14 | Output / 1M$0.58 | Context262.1K | Updated |
| ProviderDevPass (LLM Gateway) | Provider model IDhy3 | Regionglobal | Input / 1M$0.14 | Output / 1M$0.58 | Context262.1K | Updated |
| ProviderCrossModel | Provider model IDtencent/hy3 | Regionglobal | Input / 1M$0.16 | Output / 1M$0.64 | Context262.1K | Updated |
| ProviderOrcaRouter | Provider model IDtencent/hy3 | Regionglobal | Input / 1M$0.18 | Output / 1M$0.59 | Context256K | Updated |
Official prices use only the vendor's configured official Provider and positive standard USD PAYG rates. Third-party offers remain explicitly labeled.
Provider-specific output speed and catalog latency for Hy3. Runtime does not affect the capability score.
No provider-specific speed or latency record is linked to the default version yet.
Output Speed is generated output tokens received per second. Catalog latency is reported separately from observed provider TTFT.
Technical details for the model's default version.
Available versions of this model. The score column identifies the version used in the overall ranking.
Version | Released | LLMBoard | Parameters | Context | Max output | Open weights | License |
|---|
| VersionHy3 | Released | LLMBoard71.35 | Parameters295B | ContextN/A | Max outputN/A | Open weightsYes | LicenseApache 2.0 |
Recommendations prioritize the same model type and family, then the closest LLMBoard score.
Key information about Hy3 and its available data.
Hy3 is a hybrid-thinking large language model developed by the Tencent Hy Team, with configurable reasoning effort modes: no_think, low, and high. 8B MTP layer.
Data as of 2026-09-08.
Common questions about Hy3.
Hy3's default version was released on Jul 6, 2026.
No official standard PAYG price is currently available for Hy3. The lowest tracked third-party offer starts at $0.066 input and $0.26 output via NanoGPT.
Hy3 was created by Tencent.
A context window is not available for the default version.
Yes. The default version is marked as open weight under Apache 2.0.
15 provider offerings are linked to the default version.