DesignBlack Forest Labs opened FLUX 3 Image Klein: 5.6 billion parameters, runs on an 8GB card, and free for companies under 5 million dollars in annual revenueOn October 15 Black Forest Labs released FLUX 3 Image Klein with open weights: text-to-image, editing and up to 10 reference images. By the company's account the INT8 build runs on NVIDIA RTX cards with as little as 8GB of VRAM. The license is free for commercial use of outputs under 5 million dollars in annual revenue.DesignGoogle released Nano Banana 2.1 and says it does better with panoramas up to 8:1 and with characters across turnsGoogle announced on 6 October in the Gemini API release notes that the model is generally available: according to the company it keeps Flash-level speed and cost efficiency and improves the wide 1:4, 4:1, 1:8 and 8:1 formats up to 4K. The old gemini-3.1-flash-image is deprecated, with no shutdown date.DesignComfyUI launches Comfy Agent: describe what you want and it builds, runs and fixes the workflow next to youThe agent lives inside ComfyUI, knows the latest models and extensions, and works on the same canvas while you edit by hand. Available in Comfy Cloud since 1 October; the desktop version follows in a few weeks. It is a beta.DesignRunway Ads makes the ads, publishes them to Meta, Google and TikTok, and writes the next round itselfOn 30 September Runway announced Runway Ads: you connect an ad account and a brand kit, the product generates video and images, publishes the approved variants, reads the results and makes a new round based on what earned spend. For now it is piloting with select enterprise partners.DesignGemini Omni 1.1 arrived free in Google Vids for anyone with a Google accountFrom 23 September anyone with a personal or Workspace account can generate HD video scenes in Google Vids with the Omni 1.1 Flash model, without paying. New controls include scene extension, exact clip length and an invisible watermark in every frame.DesignChatGPT now draws from your doodle: Images 2.5 takes a sketch as its starting pointOn 8 September OpenAI began rolling out Images 2.5 to all ChatGPT, ChatGPT Work and Codex users. Per the company generation latency is up to 50% lower than Images 2.0, and the most interesting addition is the simplest one: you draw roughly in the chat and the model starts from the drawing.DesignRunway moved inside Premiere Pro and After Effects and saved the editor one tedious round tripOn 8 September Runway released a plugin that runs its panel inside both Adobe applications. You generate, restyle a clip from the timeline and drop it back in place without leaving. The plugin is free; generating needs a paid plan.DesignGoogle is rolling out Pics, a Nano Banana image editor that goes straight into Docs and SlidesOn 1 September Google announced Pics - an image creation and editing tool rolling out over the coming weeks to Google AI Pro and Ultra subscribers and most Workspace business customers. It works on its own and inside Docs and Slides, with Drive coming later.
17 stories