Text-to-Image
Diffusers
image-generation
korean
multilingual
hangul
korean-text-rendering
typography
poster
on-device
pocket
vidraft
Instructions to use FINAL-Bench/POCKET-Image-Zimage with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusers
How to use FINAL-Bench/POCKET-Image-Zimage with Diffusers:
pip install -U diffusers transformers accelerate
import torch from diffusers import DiffusionPipeline # switch to "mps" for apple devices pipe = DiffusionPipeline.from_pretrained("FINAL-Bench/POCKET-Image-Zimage", torch_dtype=torch.bfloat16, device_map="cuda") prompt = "Astronaut in a jungle, cold color palette, muted colors, detailed, 8k" image = pipe(prompt).images[0] - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- Draw Things
- DiffusionBee
| license: apache-2.0 | |
| base_model: | |
| - Tongyi-MAI/Z-Image | |
| pipeline_tag: text-to-image | |
| library_name: diffusers | |
| language: | |
| - ko | |
| - en | |
| - zh | |
| - ja | |
| - ar | |
| - th | |
| tags: | |
| - text-to-image | |
| - image-generation | |
| - korean | |
| - multilingual | |
| - hangul | |
| - korean-text-rendering | |
| - typography | |
| - poster | |
| - on-device | |
| - vidraft | |
| # πΌοΈ POCKET-Image-Zimage β Character-Perfect Text in Any Image | |
| **POCKET-Image** is VIDRAFT's on-device text-in-image system. It places **character-perfect text in | |
| any language** β Korean Β· δΈζ Β· ζ₯ζ¬θͺ Β· Ψ§ΩΨΉΨ±Ψ¨ΩΨ© Β· ΰΉΰΈΰΈ’ Β· Latin and more β onto any scene you describe. | |
| Built on the open, commercial-friendly [Z-Image](https://huggingface.co/Tongyi-MAI/Z-Image) foundation | |
| and VIDRAFT's **POCKET-Core** engine. | |
| > Where global models smear non-Latin scripts ("μλ νμΈμ" β "μγ κΈ°"), POCKET-Image renders every glyph exactly. | |
| ## β¨ Highlights | |
| - β **100% accurate text**, any language β RTL (Arabic) and complex shaping (Thai, Indic) handled | |
| - πΌοΈ **Any background** from a prompt β text is optional (empty β pure image) | |
| - β‘ **No GPU, no NPU** β runs on plain CPU + RAM | |
| - πͺΆ **From 4.5 GB** footprint Β· π» Windows Β· macOS Β· Linux (RTX or Apple Silicon) | |
| - π **Fully local** β no cloud, no account | |
| ## π¨ Try it | |
| - **Studio (generate here):** https://huggingface.co/spaces/FINAL-Bench/POCKET-Image-Studio | |
| ## π» On-device footprint (measured) | |
| | Machine | Peak memory | | |
| |---|---| | |
| | RTX 3050 / 4060 | **8.6 GB** | | |
| | RTX 2060 (6 GB, offloaded) | **4.5 GB** | | |
| | MacBook (M1 ProβM4, 16 GB+) | **13.4 GB** | | |
| Base model: [Tongyi-MAI/Z-Image](https://huggingface.co/Tongyi-MAI/Z-Image) (Apache-2.0). Engine: VIDRAFT POCKET-Core. | |
| <!-- POCKET-FAMILY --> | |
| --- | |
| ## π§© The POCKET Family β On-device AI by VIDRAFT | |
| *Big models, small hardware. No GPU, no cloud.* | |
| **Models** | |
| - π¦ [POCKET-35B-GGUF](https://huggingface.co/FINAL-Bench/POCKET-35B-GGUF) β flagship, PC / server, no GPU | |
| - π¦ [POCKET-26B-GGUF](https://huggingface.co/FINAL-Bench/POCKET-26B-GGUF) β compact 26B | |
| - π°π· [POCKET-KR-GGUF](https://huggingface.co/FINAL-Bench/POCKET-KR-GGUF) β Korean, Android | |
| - π [POCKET-KR-MLX](https://huggingface.co/FINAL-Bench/POCKET-KR-MLX) β Korean, iPhone / Mac | |
| - π [POCKET-EN-GGUF](https://huggingface.co/FINAL-Bench/POCKET-EN-GGUF) β English, phone / PC | |
| - πΌοΈ [POCKET-Image-Zimage](https://huggingface.co/FINAL-Bench/POCKET-Image-Zimage) β character-perfect text in any image | |
| **Demos & tools (Spaces)** | |
| - π¨ [POCKET-Image Studio](https://huggingface.co/spaces/FINAL-Bench/POCKET-Image-Studio) β text-in-image, generate in-page | |
| - π₯οΈ [POCKET-35B-CPU](https://huggingface.co/spaces/FINAL-Bench/POCKET-35B-CPU) β 35B answering on a CPU | |
| - π₯οΈ [POCKET-26B-CPU](https://huggingface.co/spaces/FINAL-Bench/POCKET-26B-CPU) β 26B on a CPU | |
| π [Full POCKET collection](https://huggingface.co/collections/FINAL-Bench/pocket-models-6a618ee5d23eafb7e185a5c6) | |
| <!-- /POCKET-FAMILY --> | |