Request a pilot
‹ PublicAI Index

The LLM benchmark aggregator.

Hy4

Tencent · 780B · open weights

Strongest in IT operations (#8 of 11), weakest in Medical (#58 of 140). Above par in 9 of 10 scopes. Among the models it meets almost everywhere, it finishes behind Claude Opus 5 and Claude Fable 5.1 and ahead of Grok 4.7 and Gemini 3.8 Flash.

§ 1 · Profile

What it is good at

Bars run from 50 — the average of the models each source lists — so right of the line is above par. The middle column is the gap to whoever leads that scope.

Coding58.1#16/164
Agentic coding59.4#15/157
Professional53.8#43/168
IT operations41.2#8/11
Biology research50.2#9/16
Legal58.5#19/151
Finance55.7#38/151
Medical53.3#58/140
Reasoning54#48/178
Mathematics56#38/140

§ 2 · Head to head

What it beats, and what beats it

The same models turn up scope after scope. Counted once: where both were placed, who finished higher. Bars run right for Hy4, left for the other.

Jamba 1.6 Miniwon 6 · lost 0
Jamba 1.6 Largewon 6 · lost 0
Mistral Small 3.1won 6 · lost 0
Llama 4 Scout Instructwon 6 · lost 0
Gemini 1.5 Prowon 6 · lost 0
Gemini 2.5 Flash Litewon 6 · lost 0
Mistral Medium 3.1won 6 · lost 0
GPT-5 Nanowon 6 · lost 0
Magistral Small 1.2won 6 · lost 0
Magistral Medium 1.2won 6 · lost 0
GPT OSS 20Bwon 6 · lost 0
Nemotron 3.5 Lightningwon 6 · lost 0
Ling 3.0 Flash Finwon 6 · lost 0
Gemma 4 31Bwon 6 · lost 0
GPT-4o Miniwon 7 · lost 0
Ling 3.0 Flashwon 7 · lost 0
Qwen3.6 27Bwon 7 · lost 0
Laguna XS.2won 7 · lost 0
Nemotron 3 Ultrawon 7 · lost 0
Laguna M.1won 7 · lost 0
GPT-5.4 Miniwon 7 · lost 0
Claude 3.5 Haikuwon 7 · lost 0
Claude 3.5 Sonnetwon 7 · lost 0
Grok 4.3won 8 · lost 0
Gemini 3.1 Flash Litewon 8 · lost 0
Command Awon 8 · lost 0
Mistral Large 3won 8 · lost 0
Mercury 2.5won 8 · lost 0
GPT-4owon 8 · lost 0
Mistral Medium 3.5won 8 · lost 0
Claude Haiku 4.5won 8 · lost 0
GPT-5.4 Nanowon 8 · lost 0
GPT-4.1 Nanowon 8 · lost 0
Qwen3.7 Maxwon 8 · lost 0
Qwen3.6 Pluswon 8 · lost 0
Gemini 3.5 Flash Litewon 8 · lost 0
Gemini 2.0 Flashwon 8 · lost 0
DeepSeek V3won 8 · lost 0
Mimo V2.5won 8 · lost 0
Inkling Smallwon 8 · lost 0
GPT-4.1 Miniwon 8 · lost 0
Mimo V2.5 Prowon 8 · lost 0
Kimi K2.6won 8 · lost 0
MiniMax M2.1won 8 · lost 0
DeepSeek V4won 8 · lost 0
Grok 3won 8 · lost 0
Claude Sonnet 4won 8 · lost 0
Gemini 2.5 Flashwon 8 · lost 0
Qwen3.8 27Bwon 8 · lost 0
Kimi K2won 8 · lost 0
DeepSeek V4 Flashwon 9 · lost 0
GLM-5.2won 9 · lost 0
GPT-6 Lunawon 8 · lost 1
GPT-5.6 Lunawon 8 · lost 1
Grok 4 Fastwon 7 · lost 1
Kimi K2.5won 7 · lost 1
Inklingwon 7 · lost 1
Kimi K2 Thinkingwon 7 · lost 1
MiniMax M2.5won 7 · lost 1
GLM-4.7won 7 · lost 1
DeepSeek V3.2won 7 · lost 1
Grok 3 Miniwon 7 · lost 1
O1won 7 · lost 1
Claude 3.7 Sonnetwon 7 · lost 1
Claude Opus 4won 7 · lost 1
O3 Miniwon 7 · lost 1
DeepSeek R1won 7 · lost 1
O4 Miniwon 7 · lost 1
GLM-4.5won 5 · lost 1
Claude Opus 4.1won 5 · lost 1
Qwen3.6 Maxwon 5 · lost 1
Gemini 3.6 Flashwon 7 · lost 2
Grok 4.1 Fastwon 6 · lost 2
Grok 4.20won 6 · lost 2
GPT OSS 120Bwon 6 · lost 2
GLM-4.6won 6 · lost 2
Qwen3 Maxwon 6 · lost 2
MiniMax M2.7won 6 · lost 2
Claude Sonnet 4.5won 6 · lost 2
Qwen3.5 Flashwon 6 · lost 2
GLM-5won 6 · lost 2
GLM-5.1won 6 · lost 2
Gemini 2.5 Prowon 6 · lost 2
Grok 4won 6 · lost 2
Mimo V2.6 Flashwon 5 · lost 2
DeepSeek V4.1 Flashwon 7 · lost 3
MiniMax M3won 5 · lost 3
GPT-5 Miniwon 5 · lost 3
Gemini 3 Flashwon 5 · lost 3
Qwen 3.5 Pluswon 5 · lost 3
Gemini 3.1 Prowon 5 · lost 3
Gemini 3.5 Flashwon 5 · lost 3
Claude Sonnet 4.6won 5 · lost 3
GLM-5.3 Flashwon 5 · lost 3
GPT-5.4won 5 · lost 3
DeepSeek V4 Prowon 5 · lost 3
Grok 4.5won 5 · lost 4
Gemini 3.8 Flashwon 5 · lost 4
Grok 4.7won 5 · lost 4
Claude Opus 4.5won 4 · lost 4
GPT-5.1won 4 · lost 4
Claude Opus 4.6won 4 · lost 4
GPT-4.1won 4 · lost 4
GPT-5won 4 · lost 4
Qwen3.8 Maxwon 4 · lost 4
Muse Spark 1.1won 4 · lost 4
Claude Sonnet 5won 4 · lost 4
GPT-5.6 Terrawon 4 · lost 4
O3won 4 · lost 4
Gemini 3.7 Flashwon 4 · lost 5
Muse Spark 1.2won 4 · lost 5
Gemini 3 Prowon 3 · lost 5
Muse Sparkwon 3 · lost 5
GPT-5.2won 3 · lost 5
Claude Opus 4.8won 3 · lost 5
Grok 4.6won 3 · lost 6
GPT-5.5won 3 · lost 6
GPT-6 Solwon 3 · lost 6
Mimo V2.6 Prowon 2 · lost 5
Claude Opus 4.7won 2 · lost 6
GLM-5.3won 2 · lost 6
Kimi K3won 2 · lost 7
GPT-6 Astrawon 2 · lost 8
GPT-5.6 Solwon 1 · lost 9
Claude Opus 5.5won 1 · lost 9
Muse Spark 1.3won 0 · lost 7
Claude Fable 5won 0 · lost 8
Claude Fable 5.1won 0 · lost 9
Claude Opus 5won 0 · lost 10

§ 3 · Sources

Where the numbers come from

1 publication, 11 figures. Every one links to the page it was read from.

Vals.ai ↗11 measuresread 2026-09-27
Vals · Legal Research Bench 45.19%Vals · LegalBench 83.76%Vals · Harvey Legal Agent Benchmark 9.17%Vals · Finance Agent 55.06%Vals · MedCode 43.25%Vals · MedScribe 83.6%Vals · BioMysteryBench 69.26%Vals · SRE Bench 2.29%Vals · Vibe Code Bench 77.48% (OpenHands)Vals · Code Migration 47.43%Vals · ProofBench 75%

Badge

PublicAI Index badge for Hy4[![PublicAI Index](https://publicai.io/model-index/badge?model=hy4)](https://publicai.io/model-index/m/hy4)