ai.canshu.com
中文
VRAM Calculator Advisor & TCO API Pricing Compare Models Leaderboard GPUs

Search models, GPUs, and tools

↑↓ Navigate · Enter Open · ★ Watchlist prioritized ai.canshu.com
Tencent 腾讯混元
Proprietary API Tencent 腾讯混元 · hy4 · Text+Image

Hunyuan Hy4 (腾讯混元 4) Official Specifications, Pricing & Benchmarks

Tencent Hunyuan 4th-gen flagship model (Hy4) with 1.05M context and 65K max output.

📋 Hunyuan Hy4 (腾讯混元 4) Official API & System Card Specifications

Verified against official Tencent 腾讯混元 developer documentation & pricing

API Model ID: hy4

1. Identity & Capabilities

API Spec
API Model ID
hy4
Release Date
2026-11-04
Weight License
Proprietary (Undisclosed)
Native Reasoning
✅ Supported
Function Calling & JSON Schema
✅ 原生支持
Batch API 异步批处理
✅ Supported (50% off)

2. Context, Output & Modalities

Limits
Max Context Window
1,049,000 tokens (1049K)
Max Output Tokens
65,536 tokens
Input Modalities
Text + Image
Output Modalities
Text
Direct China Access
✅ Direct CN + RMB
Deployment Mode
Cloud API only (Proprietary)

3. Official Pricing & Cache

USD / 1M
Input / 1M Tokens
$ 0.417
Cached Input / 1M
$ 0.021 -95%
Output / 1M Tokens
$ 1.25

🎛️ Official Sampling & Reasoning Parameters

Recommended generation settings & API parameter constraints

Official Reasoning & Tool Config
temperature=0.7 top_p=0.95 max_tokens=65,536

📊 Hunyuan Hy4 (腾讯混元 4) Full Benchmark Scorecard

32 public benchmarks · ranks among 154 models · synced 2026-11-20

⚔️ Compare head-to-head

💻 Coding 11 items

Terminal-Bench 2.1 #11/37 85.4%

Terminal-Bench 2.1 is an updated release of the Terminal-Bench benchmark that tests AI agents' ability to operate a computer via the terminal.

MCP Atlas #7/39 83.7%

MCP Atlas is a benchmark for evaluating AI models on scaled tool use capabilities, measuring how well models can coordinate and utilize multiple tools across complex multi-step tasks.

SWE-bench Multilingual #6/33 82.9%

A multilingual benchmark for issue resolving in software engineering that covers Java, TypeScript, JavaScript, Go, Rust, C, and C++.

CyberGym #9/15 78.4%

CyberGym is a benchmark for evaluating AI agents on cybersecurity tasks, testing their ability to identify vulnerabilities, perform security analysis, and complete security-related challenges in a controlled environment.

SWE-Bench Pro #6/52 65.7%

SWE-Bench Pro is an advanced version of SWE-Bench that evaluates language models on complex, real-world software engineering tasks requiring extended reasoning and multi-step problem solving.

DeepSWE #6/12 64.3%

DeepSWE is a software engineering agent benchmark evaluated with the mini-swe-agent harness, where each task is solved in an isolated container with no internet access. It measures an agent's ability to autonomously resolve real-world coding issues end to end.

SkillsBench #2/11 62.9%

SkillsBench evaluates coding agents on self-contained programming tasks, measuring practical engineering skills across diverse software development scenarios.

NL2Repo #5/23 58.9%

NL2Repo evaluates long-horizon coding capabilities including repository-level understanding, where models must generate or modify code across entire repositories from natural language specifications.

PostTrainBench #7/10 35.6%

PostTrainBench evaluates a model's ability to autonomously post-train base models. Given pretrain-only base models, the agent must complete the full pipeline of data synthesis, training, evaluation, and iteration within a time budget, scored across downstream benchmarks such as…

SWE-Marathon #6/9 31.9%

SWE-Marathon is an ultra-long-horizon software engineering benchmark covering tasks such as building compilers, optimizing kernels, and developing production-grade services. It measures whether agents can sustain quality across extremely long engineering trajectories.

Program Bench #16/18 17.5%

Program Bench evaluates code-generation agents by asking them to recreate a program's behavior from only a compiled binary and documentation.

🤖 Agents & Tools 12 items

WideSearch #1/13 83.9%

WideSearch is an agentic search benchmark that evaluates models' ability to perform broad, parallel search operations across multiple sources. It tests wide-coverage information retrieval and synthesis capabilities.

BankerToolBench #1/4 78.6%

BankerToolBench is a public benchmark that evaluates models on banking and finance tool-use tasks. Models are scored against dataset rubrics, measuring their ability to correctly invoke tools and complete multi-step financial workflows.

DRACO #2/5 77.2%

DRACO is a deep research benchmark that evaluates an agent's ability to gather, synthesize, and reason over information to answer complex research questions. Scores are based on official rubrics per question, with the final score being the average across all questions.

Toolathlon #3/32 74.1%

Tool Decathlon is a comprehensive benchmark for evaluating AI agents' ability to use multiple tools across diverse task categories. It measures proficiency in tool selection, sequencing, and execution across ten different tool-use scenarios.

OfficeQA Pro #6/15 66.2%

OfficeQA Pro evaluates AI models on professional knowledge-work questions and tasks drawn from real office workflows, including document analysis, spreadsheet reasoning, and information synthesis across business domains.

OneMillion Bench #2/3 65.4%

OneMillion Bench evaluates AI agents on high-economic-value tasks that require sustained, reliable execution across long-horizon real-world workflows.

Job Bench #4/10 61.7%

Job Bench evaluates AI agents on realistic professional tasks that require multi-step planning, research, and production of work artifacts.

Workspace Bench #1/3 60.2%

Workspace Bench evaluates AI agents on high-economic-value workplace tasks that span multi-step planning, file processing, and tool use across realistic office and productivity workflows.

GDPval-AA #6/16 1678

GDPval-AA evaluates AI agents on economically valuable professional knowledge-work tasks and reports performance as an Elo score.

APEX-Agents #8/15 37.1%

APEX-Agents is a benchmark evaluating AI agents on long horizon professional tasks that require sustained reasoning, planning, and execution across complex multi-step workflows.

AutomationBench #15/26 32.1%

AutomationBench is a tool-use benchmark that evaluates AI agents on automating real-world workflows, testing their ability to orchestrate tools and complete multi-step automation tasks.

Agents' Last Exam #28/29 22.8%

Agents' Last Exam is a challenging benchmark for AI agents on hard, long-horizon tasks that test sustained reasoning, planning, and tool use, reported with and without tool access.

Source: ZeroEval / llm-stats.com (✓ = third-party verified, others vendor-reported).

Closest Self-Hostable Open-Weight Alternatives to Hunyuan Hy4 (腾讯混元 4)

Open-weight models with full HuggingFace weights for private GPU / Mac deployment

Compare with these →

Frequently Asked Questions

Is Hunyuan Hy4 (腾讯混元 4) open-source or proprietary?

Hunyuan Hy4 (腾讯混元 4) (hy4) is a proprietary commercial model by Tencent 腾讯混元. Weights and internal layer dimensions are undisclosed; it is accessed via cloud API.

What are the context limit and API pricing of Hunyuan Hy4 (腾讯混元 4)?

Hunyuan Hy4 (腾讯混元 4) supports 1049K input context and 65,536 max output tokens, priced at $0.417/1M input ($0.021/1M cached) and $1.25/1M output.