Loading page…
Preferred languages
More languages supported by Google Translate
Open the selector to load the current language catalog.
Translations are provided by Google Translate.
Loading page…
Preferred languages
More languages supported by Google Translate
Open the selector to load the current language catalog.
Translations are provided by Google Translate.
Tencent · Open Weight · rank 26 · bench-align-v5
Hy4 preview is a general-purpose language model. It ranks in the top quarter for knowledge (top 12% of 284 scored models), by BenchLM's published scores. It accepts text and produces text.
Hy4 preview by Tencent scores 68.2 overall, costs $0.83 per 1M input tokens and $2.50 per 1M output tokens, and carries a context window of 1M tokens — observed 2026-11-21.
Seven axes from the ranking source. A missing axis is drawn as a gap.
Each axis shows the ranking source’s score and, beside it, BenchLM’s separate verified score for that axis. The two are published side by side and are not interchangeable.
Every published price component, including cache reads and writes.
Lifecycle events the source associates with this model, plus what this profile does not claim.
That is not evidence the model has no lifecycle plan — only that this source published none.
Measured values with the date they were observed. Nothing is inferred from a sibling model or a provider claim.
Time to first token: latency to the first answer chunk (Artificial Analysis, via BenchLM). Reasoning models include thinking time, so values can run to tens or hundreds of seconds.
Evidence key: ObservedUnavailable
| Measurement | Value | Observed | Last good | Evidence |
|---|---|---|---|---|
| Time to first token | Unavailable | Unavailable | Unavailable | |
| Throughput | 61.25 tok/s | 2026-11-21 | 2026-11-21 |
Regional or per-endpoint measurements appear only when the API supplies them; none are modelled here.
10 conversations per day × 8 messages × 22 active days, 1200 input and 400 output tokens per message, no cache. This uses the same calculator as the cost simulator, so an unavailable applicable rate makes the example unavailable too.
| Benchmark | Value | Rank | Tasks | Difficulty | Provenance |
|---|---|---|---|---|---|
| GPQA Graduate-Level Google-Proof Q&A | 92.3 | #21 | 448 questions | Graduate level | GPQA: A Graduate-Level Google-Proof Q&A Benchmark |
| GPQA-D GPQA Diamond | 92.3 | #23 | Graduate-level science questions | Graduate level | GPQA: A Graduate-Level Google-Proof Q&A Benchmark |
| HLE Humanity's Last Exam | 55.4 | #15 | Expert-level questions | Frontier expert level | Humanity's Last Exam |
| HLE w/o tools Humanity's Last Exam without tools | 43.4 | #23 | Expert-level questions | Frontier expert level | Introducing GPT-5.4 mini and nano |
| Benchmark | Value | Rank | Tasks | Difficulty | Provenance |
|---|---|---|---|---|---|
| Terminal-Bench 2.1 Terminal-Bench 2.1 | 85.4 | #18 | Terminal-based software-agent tasks | Professional software engineering | Terminal-Bench |
| SWE-bench Pro SWE-bench Pro | 65.7 | #14 | 1,865 repository problems | Long-horizon professional engineering | SWE-Bench Pro: Can AI Agents Solve Long-Horizon Software Engineering Tasks? |
| SWE Multilingual SWE Multilingual | 82.9 |
| Benchmark | Value | Rank | Tasks | Difficulty | Provenance |
|---|---|---|---|---|---|
| Apex Apex | 74.2 | #1 | Advanced mathematical reasoning | Frontier math reasoning | DeepSeek-V4 Technical Report |
| Benchmark | Value | Rank | Tasks | Difficulty | Provenance |
|---|---|---|---|---|---|
| CritPt Critical Physics Tasks | 16.9 | #44 | Research-level physics questions | Research-level physics reasoning | CritPt Benchmark Leaderboard |
| STEMBench STEMBench120: six-domain scientific research progress | 2.59 | #5 | 120 research tasks, 20 in each of six STEM domains | Open research targets; no run has reached a full-target success | STEMBench: six-domain model evaluation |
| Benchmark | Value | Rank | Tasks | Difficulty | Provenance |
|---|---|---|---|---|---|
| DRACO Data Research and Analysis with Complex Operations | 77.2 | #9 | Agentic data research and analysis tasks | Professional data analysis | Claude Opus 5 System Card |
| Terminal-Bench 2.1 Terminal-Bench 2.1 | 85.4 | #19 | Terminal-based software-agent tasks | Professional software engineering | Terminal-Bench |
| HLE w/ tools Humanity's Last Exam with tools | 55.4 |
| Benchmark | Value | Rank | Tasks | Difficulty | Provenance |
|---|---|---|---|---|---|
| OfficeQA Pro OfficeQA Pro | 66.2 | #6 | Document and spreadsheet tasks | Enterprise grounded reasoning | OfficeQA Pro: An Enterprise Benchmark for End-to-End Grounded Reasoning |
| Benchmark | Value | Rank | Tasks | Difficulty | Provenance |
|---|---|---|---|---|---|
| NYT Connections NYT Connections LLM Benchmark (extended version) | 68.2 | #57 | 940 NYT Connections puzzles, each with up to four added trick words | Wordplay, word knowledge and decoys that fit almost-but-not-quite | NYT Connections LLM Benchmark (Lech Mazur, GitHub) |
| StealthMark StealthMark Intelligence Benchmarks (overall score) | 43.9 | #25 | 8 original tasks: Newton's Mirror, Ground Truth, Latch, Foresight, Mona Lisa, Selfie, Beach and Raccoon | Chained physical reasoning, planning without tools and precise visual construction | StealthMark: The AI Benchmark Explained (Stealth Models) |
| #8 |
| Multilingual software-engineering tasks |
| Professional software engineering |
| MiniMax M2.7: Early Echoes of Self-Evolution |
| ProgramBench ProgramBench: Can Language Models Rebuild Programs From Scratch? | 17.5 | #20 | 200 program reconstruction tasks | Full-repository software architecture | ProgramBench: Can Language Models Rebuild Programs From Scratch? |
| PostTrainBench (lab runs) PostTrainBench lab-run results with no stated version | 35.6 | #3 | Post-training base models under the PostTrainBench budget | Autonomous ML engineering | PostTrainBench |
| NL2Repo NL2Repo | 58.9 | #8 | Natural language to repository tasks | System-level software comprehension | MiniMax M2.7: Early Echoes of Self-Evolution |
| DeepSWE DeepSWE | 64.3 | #41 | 113 software engineering tasks across 91 repositories and 5 languages | Long-horizon software engineering | DeepSWE benchmark blog |
| #19 |
| Expert questions with tool use |
| Frontier tool-augmented reasoning |
| DeepSeek-V4 Technical Report |
| GDPval-AA GDPval-AA | 1678 | #17 | Agentic real-world work tasks | Professional agentic workflows | DeepSeek-V4 Technical Report |
| CyberGym CyberGym | 78.4 | #20 | 1,507 vulnerability analysis instances | Real-world cybersecurity | CyberGym: Evaluating AI Agents' Real-World Cybersecurity Capabilities at Scale |
| JobBench JobBench | 61.7 | #9 | 130 tasks across 35 occupations | Professional multi-source workflows | JobBench: Aligning Agent Work With Human Will |
| MCP Atlas MCP Atlas | 83.7 | #8 | Tool-integrated agent tasks | Advanced tool use | Introducing GPT-5.4 mini and nano |
| Toolathlon-Verified Toolathlon-Verified | 74.1 | #8 | Verified multi-tool workflows | Advanced tool use | Kimi K3: Open Frontier Intelligence |
| AutomationBench AutomationBench | 32.1 | #25 | 600 public automation tasks | Long-horizon automation | Kimi K3: Open Frontier Intelligence |
| Agents' Last Exam Agents' Last Exam | 22.8 | #22 | Agent tasks | Advanced agentic work | DeepSeek V4 Flash 0731 update |
| APEX-Agents APEX-Agents | 37.1 | #9 | Professional-services agent tasks | Long-horizon professional work | Kimi K3: Open Frontier Intelligence |
| WideResearch WideResearch | 83.9 | #1 | Open-ended research tasks | Broad research-agent workflows | Qwen3.6 launch benchmarks |
| BankerToolBench BankerToolBench | 78.6 | #1 | Finance and banking tool-use tasks | Professional finance-agent workflows | MiniMax M3 model card |
| Terminal-Bench 2.1 (Vals) Terminal-Bench 2.1, Vals AI run | 55.1 | #45 | Difficult terminal tasks | Frontier agentic | Vals AI Terminal-Bench 2.1, Vals AI run leaderboard |
| CWE-bench v1 CWE-bench v1 | 53.0 | #10 | Audit and patch held-out C/C++, Go, Java, TypeScript/JavaScript, Python, and Rust repositories | Frontier agent evaluation | CWE-bench v1 leaderboard and methodology |
| Hermes Index Hermes Index: agentic capability and cost per task inside Hermes Agent | 41.0 | #6 | Four suites: Hermes Bench (150 tasks), TerminalBench 4 without its 4 GPU tasks, TerminalBench Science and SkillsBench | Frontier agentic | Hermes Index: an agentic model leaderboard |
| Hermes Bench Hermes Bench: 150 agent tasks graded on the files and state the agent leaves behind | 70.8 | #10 | 150 tasks across 25 categories | Frontier agentic | Hermes Bench (Hermes Index) |
| CWE-bench CWE-bench v2 | 47.3 | #16 | Audit and patch C/C++, Go, Java, TypeScript/JavaScript, Python, and Rust repositories; 100 public and 150 held-out tasks | Frontier agent evaluation | CWE-bench v2 leaderboard and methodology |