Qwen Image 3.0 and Qwen Image 3.0 Pro are hosted API models, announced on 22 July 2026 and generally available on Alibaba Cloud Model Studio since 4 August 2026, with no model card, no technical report and no published parameter count, according to the article. The open lineage stops one step earlier, at Qwen-Image-2.1, released in September 2026 as a 7B diffusion transformer paired with a Qwen3-VL 8B text encoder under the Qwen Research License — not Apache 2.0, and not free for commercial use without a separate arrangement. Images 2.6 — OpenAI's 4 November 2026 release of GPT-Image-2.6 Sunburst and GPT-Image-2.6 Flare — is closed. The article states that aggregator pages listing Qwen Image 3.0 as open-weight under Apache 2.0 or a "Qwen License" appear to be template artefacts, and that no Hugging Face repository, no safetensors checkpoint and no model card for 3.0 or 3.0 Pro could be located. It says Qwen Image 3.0 is better than 2.1 on every measured axis but cannot be self-hosted, cannot be fine-tuned on your own data, and cannot be pinned to a checkpoint you control. Artificial Analysis, 5 November 2026, both