Qwen-Image-2512 (20B) + Edit-2511

Strongest open-weight image model for text rendering — Arabic, Chinese, English all sharp. Qwen-Image-2512 (Dec 31 2025) is the latest released generation model and claims #1 open-source on AI Arena; pair with Qwen-Image-Edit-2511 (Dec 23 2025) for editing workflows. A unified 7B "Qwen-Image 2.0" was announced Feb 10 2026 (arxiv 2605.10730 tech report) but the weights are not yet open-sourced — Qwen-Image-2512 remains the runnable flagship as of June 2026.

License: Apache 2.0 · Context: n/a (diffusion); native 2K resolution · Released: December 31, 2025 (Qwen-Image-2512, latest gen); December 23, 2025 (Qwen-Image-Edit-2511); August 4, 2025 (original 20B base)

The decision in five lines

The call: Buy — for image
Best for: image
Runs on: 21 hardware picks fit (cheapest: Minisforum UM890 Pro · $463)
Watch out: Speed-sensitive batch generation where text rendering isn't the priority — Z-Image-Turbo is ~3× faster at comparable quality on short prompts.
Evidence: Estimated · last verified June 2026

20B MMDiT (original Qwen-Image lineage): PARAMETERS
IMAGE GEN + DEDICATED EDITOR: TYPE
n/a: CONTEXT
~12 GB FP8 / ~20 GB bf16 (20B class): VRAM AT Q4

Where we recommend this

Every tier slot in the planner where this model is a top or alternate pick. Pulled live from planner.js — when the planner refreshes, this table stays current.

IMAGE ·

Qwen-Image-2512 (20B) + Edit-2511Latest released Qwen-Image gen model (Dec 31 2025) + the Dec 23 2025 editor — both Apache 2.0, both 20B, best-in-class text rendering (CN + EN). Frontier hardware keeps both checkpoints resident for fast iteration. Qwen-Image 2.0 (the announced 7B unified successor) is still weights-pending as of June 2026.

IMAGE · TOP

Qwen-Image-2512 (20B, Apache 2.0)Latest released Qwen-Image (Dec 31 2025); claims #1 open-source on AI Arena. Best-in-class text rendering (CN + EN); pair with Edit-2511 for editing workflows.

IMAGE · HIGH

Qwen-Image-2512 (20B, Apache 2.0)Dec 31 2025 release; latest runnable Qwen-Image. Best-in-class CN + EN text rendering at this tier; Apache 2.0 — rare at this quality.

The call

Strongest open-weight image model for text rendering — Arabic, Chinese, English all sharp. Qwen-Image-2512 (Dec 31 2025) is the latest released generation model and claims #1 open-source on AI Arena; pair with Qwen-Image-Edit-2511 (Dec 23 2025) for editing workflows. A unified 7B "Qwen-Image 2.0" was announced Feb 10 2026 (arxiv 2605.10730 tech report) but the weights are not yet open-sourced — Qwen-Image-2512 remains the runnable flagship as of June 2026.
When not to use: Speed-sensitive batch generation where text rendering isn't the priority — Z-Image-Turbo is ~3× faster at comparable quality on short prompts. FLUX.2 klein 4B if Apache-2.0 with sub-second iteration matters more than text fidelity.

Runner notes

ComfyUI native workflow; GGUF + FP8 quants from unsloth and vantagewithai. Two-checkpoint pipeline (gen + edit) until Qwen-Image 2.0 weights ship. Watch the official QwenLM/Qwen-Image GitHub README for the 2.0 release announcement — qwenimages.com is a third-party marketing blog, not an Alibaba property.

License: Apache 2.0
Released: December 31, 2025 (Qwen-Image-2512, latest gen); December 23, 2025 (Qwen-Image-Edit-2511); August 4, 2025 (original 20B base)
Maker: Alibaba
Model card: huggingface.co/Qwen/Qwen-Image-2512 →

Hardware that fits

Every hardware pick whose memory fits this model at the quant we recommend. Sorted cheapest-first — the top row is your best-value fit. Click through for the full buyer’s guide.

Next step

Find-by-model — see what hardware runs this→