Qwen releases Image 2.1, a new image generation model

Qwen released Image 2.1, a new image generation model, according to a post at qwen.ai on September 20, 2026. The announcement was surfaced via Hacker News.

The source material provided to this desk consists of the title, the originating URL, and publication metadata. No summary text accompanied the release notice, and no further description of the model’s capabilities, architecture, or training was included in the record available here.

The listing identifies the work as “Qwen Image 2.1,” distinguishing it from an unspecified predecessor by version number, which implies at least one prior release in the same line, though the supplied material does not describe that earlier version or confirm its existence beyond the naming convention.

No pricing, availability, licensing terms, or technical benchmarks were included in the material provided. No statements from Qwen representatives were quoted in the source record, and none are reproduced here.

This bulletin records the release notice as given. Readers seeking capability comparisons, sample outputs, or technical specification should consult the linked announcement directly, as the Specola’s log reflects only what the referenced posting and its metadata state.