DataLearner logo
HY

Hy3

Reasoning modelCoding modelHunyuan

Tencent Hy3

Release date: 2026-07-06Updated: 2026-07-17 21:57:45.7051,213
Live demoGitHubHugging FaceCompare
Parameters
295B
Context length
256K
Chinese support
Supported
Reasoning ability

Hy3 is a reasoning model from Tencent AI Lab, released on 2026-07-06. It accepts text input and returns text output. The cataloged parameter count is 295B, with 21B active parameters per inference. The recorded context window is 256K. Cataloged capabilities include Reasoning model, Multilingual, and Coding model. The checkpoint is listed under the Apache 2.0 license, with the license link included in the model references. Use the linked references to confirm current access, licensing, and provider-specific limits.

Data sourced primarily from official releases (GitHub, Hugging Face, papers), then benchmark leaderboards, then third-party evaluators. Learn about our data methodology

Hy3

Model basics

Reasoning traces
Supported
Thinking modes
Standard Mode (Default)Thinking Level · LowThinking Level · High
Context length
256K tokens
Max output length
No data
Model type
Reasoning model
Modality (in / out)
Text → Text
Release date
2026-07-06
Model file size
No data
MoE architecture
Yes
Total params / Active params
295B / 21B
Knowledge cutoff
No data
Hy3

Open source & experience

Code license
Apache 2.0
Weights license
Apache 2.0- Commercial use permitted
GitHub repo
https://github.com/Tencent-Hunyuan/Hy3
Hugging Face
https://huggingface.co/tencent/Hy3
Live demo
https://aistudio.tencent.com/
Hy3

Official resources

Paper
Paper
DataLearnerAI blog
No blog post yet
Hy3

API details

API speed
3/5
💡Default unit: $/1M tokens. If vendors use other units, follow their published pricing.
Learn about pricing modes
Standard
TypeConditionInputOutput
TextContext < 16000¥1.20/ 1M¥4.00/ 1M
TextContext >= 16000, Context < 32000¥1.60/ 1M¥6.40/ 1M
TextContext >= 32000¥2.00/ 1M¥8.00/ 1M
Cache PricingPrompt Cache
TypeTTLWriteRead
Text--¥0.400/ 1M
Context < 16000
Text--¥0.600/ 1M
Context >= 16000, Context < 32000
Text--¥0.800/ 1M
Context >= 32000
Hy3

Benchmark Results

Hy3 currently shows benchmark results led by AA-LCR (1 / 14, score 73.40), HLE (15 / 170, score 53.20), IMO-AnswerBench (2 / 21, score 90). This page also consolidates core specs, context limits, and API pricing so you can evaluate the model from benchmark results and deployment constraints together.

Thinking
Tool usage
Internet

General Knowledge

3 evaluations
Benchmark / mode
Score
Rank/total
GPQA Diamond
High
90.40
19 / 187
HLE
High
37
69 / 170
HLE
HighTools
53.20
15 / 170

Coding and Software Engineer

4 evaluations
Benchmark / mode
Score
Rank/total
SWE-bench Verified
HighTools
78
23 / 111
SWE-bench Multilingual
HighTools
75.80
6 / 22
SWE-Bench Pro - Public
HighTools
57.90
14 / 51
DeepSWE
HighTools
28
16 / 17

AI Agent - Information Search

1 evaluations
Benchmark / mode
Score
Rank/total
BrowseComp
HighToolsInternet
84.20
9 / 52

AI Agent - Tool Usage

3 evaluations
Benchmark / mode
Score
Rank/total
MCP-Atlas
HighTools
79.10
7 / 27
TerminalBench 2.1
HighTools
71.70
15 / 25
Tool Decathlon
HighTools
48.50
3 / 9

Math and Reasoning

1 evaluations
Benchmark / mode
Score
Rank/total
IMO-AnswerBench
High
90
2 / 21

Long Context

1 evaluations
Benchmark / mode
Score
Rank/total
AA-LCR
High
73.40
1 / 14
View benchmark analysisCompare with other models

Compare with other models

No curated comparisons for this model yet.

Want a custom combination? Open the compare tool

Hy3

Publisher

Tencent Hy3

Model Overview

Hy3 is a reasoning model from Tencent AI Lab, released on 2026-07-06.

It accepts text input and produces text output. Its cataloged capabilities include Reasoning model, Multilingual, and Coding model. The cataloged parameter count is 295B, with 21B active parameters per inference. The recorded context window is 256K.

The checkpoint is listed under the Apache 2.0 license, with the license link included in the model references. The page records 9 API pricing rules from the listed provider; current provider pricing and conditions should be checked before deployment. The evaluation section contains 13 cataloged benchmark results with their recorded modes and scores. The page links 4 release, model-card, repository, or provider references for checking the underlying claims. Specifications, availability, and prices can change; undisclosed values are intentionally left unstated.

Hy3

FAQ

What is Hy3?

Hy3 is a reasoning model from Tencent AI Lab, released on 2026-07-06. It accepts text input and returns text output. The cataloged parameter count is 295B, with 21B active parameters per inference. The recorded context window is 256K. Cataloged capabilities include Reasoning model, Multilingual, and Coding model. The checkpoint is listed under the Apache 2.0 license, with the license link included in the model references. Use the linked references to confirm current access, licensing, and provider-specific limits.

What input and output modalities does Hy3 support?

The current model record lists text as input and text as output.

What are the main recorded specifications for Hy3?

The cataloged parameter count is 295B, with 21B active parameters per inference. The recorded context window is 256K. Fields without a source-backed value remain undisclosed.

Does Hy3 have API pricing?

The page records 9 API pricing rules from the listed provider; current provider pricing and conditions should be checked before deployment.

Are benchmark results available for Hy3?

The evaluation section contains 13 cataloged benchmark results with their recorded modes and scores. Compare only results that use the same benchmark version and evaluation mode.

Is Hy3 open source?

The checkpoint is listed under the Apache 2.0 license, with the license link included in the model references. Review the linked license text before commercial or derivative use.

DataLearner on WeChat

Follow DataLearner on WeChat for AI model updates and research notes.

DataLearner WeChat QR code