AI Tools Review

Tencent Hy4

By Tencent

Released: 2026-11-03

LLM
Tencent
Open Weights
Free
New

Tencent Hy4 is Tencent's open-weight AI model, released 03/11/2026. Tencent's Hy4 model card publishes few-shot results for the Hy4 base model as text and shows the instruct-model benchmark charts only as images, which could not be read. The text-readable official figures are the base-model table (Hy4 Base leads Hy4 flash Base, Hy3 preview Base and both DeepSeek V4 base models on most rows), a 579-task finance evaluation where Hy4 scores 3.04 out of 4, and WorkIndex 51.4. The WorkIndex sub-tests come from our Hy4 flash article citing Tencent's flash card. Instruct-model coding and agent benchmark rows are not included because their values could not be read. The release date is the Hugging Face repository creation date.

Visit Tencent Hy4

AI-Powered

Leverages advanced AI technology to deliver cutting-edge capabilities and results.

Fast & Efficient

Optimized performance ensures quick results without compromising on quality.

Purpose-Built

Specifically designed for llms tasks and workflows.

How does Tencent Hy4 score on independent benchmarks?

Tencent Hy4 ranks #17 of 28 models on the Artificial Analysis Intelligence Index (v5.0.1) with a score of 34.6, as measured on 22 November 2026.

Intelligence Index
34.6
AA v5.0.1
Coding Agent Index
n/a
Cost per task
$1.03
AA Intelligence Index task
Price per 1M tokens
$0.834 / $2.501
input / output

Where does Tencent Hy4 rank against other models?

How does Tencent Hy4 do on individual tests?

Evaluation (run by Artificial Analysis)Tencent Hy4Best of any model
Terminal-Bench 4.037.4%67.2%
FrontierCode 2.031.4%48.8%
Humanity's Last Examn/a74.3%
GDPval-AA (Elo)n/a1866

How has Tencent Hy4's score changed over time?

We have tracked Tencent Hy4 on Intelligence Index v5.0.1 since 22/11/2026. A point is added each day the score, coding index or price changes, so this chart fills in over time.

What does the Tencent Hy4 system card report?

Tencent's Hy4 model card publishes few-shot results for the Hy4 base model as text and shows the instruct-model benchmark charts only as images, which could not be read. The text-readable official figures are the base-model table (Hy4 Base leads Hy4 flash Base, Hy3 preview Base and both DeepSeek V4 base models on most rows), a 579-task finance evaluation where Hy4 scores 3.04 out of 4, and WorkIndex 51.4. The WorkIndex sub-tests come from our Hy4 flash article citing Tencent's flash card. Instruct-model coding and agent benchmark rows are not included because their values could not be read. The release date is the Hugging Face repository creation date.

These are Tencent's own published results (self-reported, with the vendor's harness and settings), so they are not directly comparable with other labs' numbers. Sources: Hy4 (Hugging Face model card) (29/10/2026), Hy4 flash (Hugging Face model card) (16/11/2026). Last checked 22/11/2026.

Knowledge & science

BenchmarkTencent Hy4Hy4 flash BaseHy3 preview BaseDeepSeek-V4-flash-Base
MMLU (Base)5-shot; macro-averaged accuracy; base model91.6%88.7%87.5%89.7%
MMLU-Pro (Base)5-shot; macro-averaged accuracy; base model73.9%71.1%65.8%68.6%
MMLU-Redux (Base)5-shot; macro-averaged accuracy; base model90.6%87.6%87%88.8%
SuperGPQA (Base)5-shot; macro-averaged accuracy; base model62.7%57.8%52.3%52.5%
C-Eval (Base)5-shot; macro-averaged accuracy; base model95%92%90.1%91.9%
CMMLU (Base)5-shot; macro-averaged accuracy; base model93.1%91.6%89.6%89.8%
SimpleQA (Base)5-shot; base model57.2%37.5%26.4%31.3%
Chinese-SimpleQA (Base)5-shot; macro-averaged accuracy; base model83.9%74.6%69.6%71.9%
AGIEval English (Base)3-shot; macro-averaged accuracy; base model79.3%75.4%76%76.3%
AGIEval Chinese (Base)3-shot; macro-averaged accuracy; base model91.2%90.6%87.8%88.5%

Reasoning & maths

BenchmarkTencent Hy4Hy4 flash BaseHy3 preview BaseDeepSeek-V4-flash-Base
HellaSwag (Base)10-shot; 3,000-item subset; base model87.1%85.2%85.7%85.1%
PIQA (Base)4-shot; base model85.3%84.4%84.6%83.8%
ARC-Challenge (Base)0-shot; 299-item subset; base model97.3%95%96%97.3%
Social IQa (Base)0-shot; base model82.7%80.1%80.5%72.5%
KOR-Bench (Base)3-shot; base model57.6%54.7%54.6%53.4%
C3 (Base)0-shot; base model82.4%77.3%75.9%74.1%
U-MATH (Base)5-shot; base model71.6%69.8%60.9%57.4%
OCW Courses (Base)5-shot; base model66.7%64%62.6%63.3%

Coding

BenchmarkTencent Hy4Hy4 flash BaseHy3 preview BaseDeepSeek-V4-flash-Base
MBPP-plus (Base)3-shot; base model74%73.7%77.4%76.8%

Other

BenchmarkTencent Hy4Hy4 flash BaseHy3 preview BaseDeepSeek-V4-flash-Base
Belebele, 5-language subset (Base)5-shot; macro-averaged accuracy; 4,475 examples; base model95%93.8%94.2%94.4%
INCLUDE, 5,000-example Lite subset (Base)5-shot; macro-averaged accuracy; 44 languages; base model88.1%82.8%84.3%84%
Finance workflow evaluation (instruct)579 real-world finance tasks, average score out of 4; Hy4 instruct model3.04 /4n/an/an/a
WorkIndex (Tencent internal)Tencent internal set of eight office-task tests, run in WorkBuddy; figure from the Hy4 flash model card51.4 scoren/an/an/a

Multimodal

BenchmarkTencent Hy4Hy4 flash BaseHy3 preview BaseDeepSeek-V4-flash-Base
MMMU-Pro Standard, 10 options (Base)4-shot; base model59.6%52%n/an/a
SimpleVQA (Base)4-shot; base model54.5%48.5%n/an/a
CVBench (Base)4-shot; base model86.7%83.4%n/an/a
RealWorldQA (Base)4-shot; base model72.3%71.1%n/an/a
TextVQA (Base)4-shot; base model78.2%73.4%n/an/a
CountBenchQA (Base)4-shot; base model90.6%86.6%n/an/a

Agents & tool use

BenchmarkTencent Hy4Hy4 flash BaseHy3 preview BaseDeepSeek-V4-flash-Base
PPTGenBench (slide creation)WorkIndex test; run in the WorkBuddy command-line tool61.5 scoren/an/an/a
PPTCoEditBench (slide editing)WorkIndex test; run in the WorkBuddy command-line tool78.9 scoren/an/an/a
PPTFixBench (slide repair)WorkIndex test; run in the WorkBuddy command-line tool45.6 scoren/an/an/a
MdCoEditBench (Markdown co-editing)WorkIndex test; run in the WorkBuddy command-line tool57.7 scoren/an/an/a
WordEditBench (document editing)WorkIndex test; run in the WorkBuddy command-line tool57.2 scoren/an/an/a
ExcelFixBench (spreadsheet repair)WorkIndex test; run in the WorkBuddy command-line tool27.8 scoren/an/an/a
AudioVideoBenchWorkIndex test; run in the WorkBuddy command-line tool27.4 scoren/an/an/a
BuddyVQA Atlas (image questions)WorkIndex test; run in the WorkBuddy command-line tool55.5 scoren/an/an/a

What did safety testing find for Tencent Hy4?

Safety level: No vendor safety rating on the model card

  • The Hy4 model card text lists benchmarks, licence and inference settings; no safety evaluation was found. The instruct benchmark images, including a limitations chart, could not be read.

Independent figures: Artificial Analysis, refreshed daily on our AI model benchmarks page.

Tencent Model Timeline

Tencent Hy4Current

1,048,576 tokens (1M) context

Tencent: Hunyuan A13B Instruct

131k tokens context

Specifications

pricing$0.834/$2.501 per 1M tokens (input/output)
context Window1,048,576 tokens (1M)