FreeLLMAll models › tencent
HunyuanImage 3.0
tencent · tencent-hunyuan-community license · on Arena since 1 Oct 2025
Lab chat Download
Arena score1151#33 of 82 in Image, paid models included
Size83 Bparameters
4-bit download50 GBweights only, Q4_K_M
Max context22Ktokens, from the model config
Where to use it for free
- The lab's own chat: yuanbao.tencent.com (free with daily limits)
- Download and run it at home: official weights · GGUF files for Ollama and LM Studio
Arena scores
| Arena | Score | Rank | Votes |
|---|---|---|---|
| Image | 1151 | #33 of 82 | 180K |
| Image edit | 1302 | #29 of 60 | 331K |
Rank among all models, paid ones included. Source: Arena leaderboard dataset.
Does it fit on popular hardware?
| Hardware | GPU memory | Fits? |
|---|---|---|
| RTX 3060 12 GB | 12 GB | GPU + 64 GB RAM |
| RTX 4060 Ti 16 GB | 16 GB | GPU + 64 GB RAM |
| RTX 3090 | 24 GB | GPU + 64 GB RAM |
| RTX 4090 | 24 GB | GPU + 64 GB RAM |
| RTX 5090 | 32 GB | GPU + 64 GB RAM |
| RX 7900 XTX | 24 GB | GPU + 64 GB RAM |
| 2 × RTX 3090 | 48 GB | GPU + 64 GB RAM |
| 4 × RTX 3090 | 96 GB | Fits in GPU |
| Mac M4 Max 128 GB | 96 GB | Fits in GPU |
| Ryzen AI Max+ 395 128 GB | 96 GB | Fits in GPU |
| 8 × H100 SXM | 640 GB | Fits in GPU |
Estimate at 4-bit, 64 GB of DDR5 for offload. Speed ≈ memory bandwidth × 78% ÷ data read per token, plus a fixed cost per token, calibrated on 44 real llama.cpp tests; typical error about ±10%. Compare all GPUs.
GPU memory needed
| Level | Total |
|---|---|
| Q3_K_M | 45 GB |
| Q4_K_M | 54 GB |
| Q5_K_M | 63 GB |
| Q6_K | 72 GB |
| Q8_0 | 92 GB |
| F16 | 170 GB |
Weights + text encoders and decoder (about 3 GB) + runtime. Simulate your own setup.
Image score over time
120013001400
JunJulAugSeptOct
HunyuanImage 3.0 1151Best paid model
Arena score from 5 May 2026 to 26 Oct 2026.
Image edit score over time
130014001500
JunJulAugSeptOct
HunyuanImage 3.0 1302Best paid model
Arena score from 5 May 2026 to 26 Oct 2026.
Similar free models in Image
- Cosmos3-Super-Text2Image 1151 · 64.6 B
- FLUX.2 [dev] 1145
- Cosmos3-Super-Text2Image (Agentic) 1164
- Qwen-Image (2512) 1125
- HiDream-O1-Image 1115 · 8.8 B
- Krea 2 Turbo 1114 · 12.8 B
Share this page:XLinkedInRedditWhatsApp
Spotted a wrong number or a broken link? Report a mistake