Hy-MT2 is a family of “fast-thinking” multilingual translation models designed for complex real-world scenarios. It includes three model sizes: 1.8B, 7B, and 30B-A3B (MoE), all of which support translation among 33 languages and effectively follow translation instructions in multiple languages.
plamo-3-translate-8b
pfnet·Base model·released 27 Oct 2026
ollama run hf.co/pfnet/plamo-3-translate-8b-GGUF:Q4_K_MPLaMo 3 Translate 8B is a translation-specialized LLM developed by Preferred Networks (PFN) that delivers world-class translation quality.
How much VRAM does plamo-3-translate-8b need?
At Q8_0 plamo-3-translate-8b needs about 11 GB of GPU memory.
| Quantisation | Total |
|---|---|
| 4-bitRecommended default for local use | ≈6.4 GB |
| 8-bitrecommendedLossless in practice, twice the size of Q4 | ≈11 GB |
| FP16 / BF16Full precision, as released | ≈20 GB |
Estimates for the weights plus runtime overhead; activations vary with resolution and batch size. Pick a GPU above and the fit column checks every row against it. Weights marked “file” use the real GGUF download size; “est.” rows are computed from the parameter count.
Run plamo-3-translate-8b locally
--quantQ4_K_Mdefault; pick your GPU to tailor itpfnet/plamo-3-translate-8b-GGUF
ollama run hf.co/pfnet/plamo-3-translate-8b-GGUF:Q4_K_M llama-server -hf pfnet/plamo-3-translate-8b-GGUF:Q4_K_MDownloads the GGUF on first run. Installed through llama.app? The same flags work with `llama serve`.
lms get pfnet/plamo-3-translate-8b-GGUF@Q4_K_MThen load it from the LM Studio app or with `lms load`.
vllm serve pfnet/plamo-3-translate-8b vllm serve pfnet/plamo-3-translate-8b-FP8services:
vllm:
image: vllm/vllm-openai:latest
command: pfnet/plamo-3-translate-8b
ports:
- "8000:8000"
ipc: host
volumes:
- ~/.cache/huggingface:/root/.cache/huggingface
deploy:
resources:
reservations:
devices:
- driver: nvidia
count: all
capabilities: [gpu]API on port 8000.
from transformers import pipeline
pipe = pipeline("translation", model="pfnet/plamo-3-translate-8b")
result = pipe("Self-hosting puts you in control of your data.")
print(result)Quantised versions and downloads
| Repository | Format | Quant | Size |
|---|---|---|---|
| pfnet/plamo-3-translate-8b-W4A16 | Safetensors | INT4 | — |
| pfnet/plamo-3-translate-8b-NVFP4 | Safetensors | NVFP4 | — |
| pfnet/plamo-3-translate-8b-FP8 | Safetensors | FP8 | — |
| mradermacher/plamo-3-translate-8b-i1-GGUF | GGUF | IQ1_S | 2.3 GB |
| mradermacher/plamo-3-translate-8b-i1-GGUF | GGUF | IQ1_M | 2.4 GB |
| mradermacher/plamo-3-translate-8b-i1-GGUF | GGUF | IQ2_XXS | 2.6 GB |
| mradermacher/plamo-3-translate-8b-i1-GGUF | GGUF | IQ2_XS | 2.9 GB |
| mradermacher/plamo-3-translate-8b-i1-GGUF | GGUF | IQ2_S | 3.0 GB |
| mradermacher/plamo-3-translate-8b-i1-GGUF | GGUF | IQ2_M | 3.2 GB |
| mradermacher/plamo-3-translate-8b-i1-GGUF | GGUF | Q2_K_S | 3.2 GB |
About plamo-3-translate-8b
PLaMo 3 Translate 8B is a translation-specialized LLM developed by Preferred Networks (PFN) that delivers world-class translation quality. It is built on PLaMo 3 NICT, a pretrained model developed through joint research by PFN and the National Institute of Information and Communications Technology (NICT). We further trained this base model on tens of billions of tokens of translation-focused data. For development details, see the tech blog. PLaMo 3 Translate 8B supports translation among 53 languages (see Supported Languages). It can also take surrounding context and a glossary into account, and lets you choose between plain and polite style for Japanese output. The maximum length during training was 32,768 tokens for the input and output combined. The model can translate long passages in one go, but split anything longer so that the input and output together stay within this limit. PLaMo 3 Translate 8B is available under the PLaMo Community License. Personal use, such as learning or hobby projects, and other non-commercial use, including academic use, are free of charge with no prior registration. See License for the terms of commercial use.
Full model card on Hugging Face →Other Plamo 3 Translate models
Whole family →Alternatives to plamo-3-translate-8b
All alternatives →Index-Translate-9B is the 9B-parameter multilingual text model of the Index-Translate family, built on Qwen3.5. It supports text translation across 150 languages and follows translation instructions involving terminology, formatting, style, structure, context, and output length.
Hunyuan-MT-Chimera-7B-fp8 was produced by AngelSlim. The Hunyuan Translation Model comprises a translation model, Hunyuan-MT-7B, and an ensemble model, Hunyuan-MT-Chimera.
We are excited to announce our official partnership with WMT26. We welcome all participants to use our HY-MT model during the competition. Teams that use HY-MT and achieve notable results will be eligible for cash prizes. For more details, please contact us at hunyuan@tencent.com.
We are excited to introduce Seed-X, a powerful series of open-source multilingual translation language models, including an instruction model, a reinforcement learning model, and a reward model. It pushes the boundaries of translation capabilities within 7 billion parameters.
PLaMo Translation Model is a specialized large-scale language model developed by Preferred Networks for translation tasks. For details, please refer to the blog post and press release.
plamo-3-translate-8b FAQ
- How much VRAM does plamo-3-translate-8b need?
- About 11 GB at Q8_0. Full precision needs about 20 GB.
- Can plamo-3-translate-8b run on the RTX 4060?
- Yes. It fits at Q4_K_M (about 6.4 GB). It fits comfortably on the Apple M4 (16 GB), Radeon RX 6700 XT, RTX 3060 12GB.
- Can I use plamo-3-translate-8b commercially?
- Yes, with conditions. It is released under plamo-community-license. Custom licence: read the model card before using it commercially.
- How do I run plamo-3-translate-8b with Ollama or llama.cpp?
- With Ollama: ollama run hf.co/pfnet/plamo-3-translate-8b-GGUF:Q4_K_M. With llama.cpp: llama-server -hf pfnet/plamo-3-translate-8b-GGUF:Q4_K_M, which starts an OpenAI-compatible server.
Rate this model
Data from Hugging Face, the model card and public leaderboards. Memory and speed figures are estimates; real numbers depend on the runtime, context length and drivers. Always check the licence yourself before commercial use.
0 comments