Skip to main content

The news desk

AI news, checked against the source.

What changed in AI, one story at a time. Each one links to its primary source, and facts stay apart from interpretation.

Last updated

Friday, October 2UTC

  1. New
    MidjourneyMidjourney Alpha workflow updateImages

    Midjourney Alpha fixes edit proportions, style settings and search recovery

    Midjourney Alpha now keeps the base image aspect ratio during edits even when a stray aspect parameter is present. Sending a style preview preserves prompt settings, folder parameters persist within a tab session, and Back restores the previous feed page after search. Organization selection and several narrow-screen controls also receive fixes. Collaborative tools and persistent edit history are listed as upcoming, not shipped features.

    Midjourney1 source

  2. New
    RunwayIdeogram 4.5 in Runway MCPImages

    Runway MCP adds Ideogram 4.5 for paid plans

    Paid-plan MCP workflows can now use Ideogram 4.5 through Runway. This integration is separate from the original model release.

    Runway2 sources

  3. New
    QwenQwen-Image-2.1-ProImages

    Alibaba Cloud lists Qwen-Image 2.1 Pro for international API use

    Alibaba Cloud's model-release log lists Qwen-Image-2.1-Pro for its international service on October 2. The entry combines image creation and editing, native transparency and up to ten reference images. Its stated 7-billion-parameter figure applies to the visual-generation component, not necessarily the entire system. This hosted availability event is separate from the September 20 open-weight research release, and does not establish new unrestricted weights or universal regional access.

    Alibaba Cloud1 source

Free, email only

Unlock the full brief free.

  • Thirty days of stories to browse, not seven
  • On every story: what to check, each source and why it matters
  • The same signup unlocks every free tool

This is not an account: there is no password, and the unlock is a cookie in this browser. You also join the Rise Productive newsletter from Demetri Panici, about once a week: what I built and what changed in AI. We'll email you a link to confirm, and you can unsubscribe in one click. The same signup unlocks every free tool on the site. How your email is handled.

Already subscribed? Enter the same email to unlock this browser. You won't be signed up twice.

Thursday, October 1UTC

  1. New
    HiggsfieldFLUX 3 Image on HiggsfieldImages

    Higgsfield adds FLUX 3 Image generation and editing

    Higgsfield's October 1 changelog adds FLUX 3 Image for text-to-image and image-to-image work, with output choices through 4K. This is provider availability, not the original Black Forest Labs model launch.

    Higgsfield1 source

  2. New
    Black Forest LabsFLUX 3 ImageImages

    FLUX 3 Image adds box-guided editing, ten references and up-to-4K output

    BFL's October 1 announcement and live API docs establish image generation and targeted editing with up to ten references. Bounding boxes guide composition. Pixel-preservation claims are the publisher's, not an independent benchmark; a separate open-weights edition is promised rather than available now.

    Black Forest Labs3 sources

  3. New
    Black Forest LabsFLUX 3 Image API promotionImages

    BFL offers 50% off FLUX 3 Image API use through October 8

    The original BFL reply announces a time-limited API promotion, and the current default pricing calculator displays a discounted rate. This is separate from the model launch and is not a permanent pricing cut or a discount on every resolution and licensing option without checking the applicable terms.

    Black Forest Labs2 sources

  4. New
    LumaAd VariantsImages

    Luma Variants adapts one approved static ad across sizes and languages

    Luma announced Ad Variants for adapting an approved static master into format-and-language combinations. Five placement presets include story, portrait, square, rectangle and widescreen. Users preview and regenerate results, then download them or export to a Luma Canvas board. The first release covers static ads, resizing and translation; it does not promise video variants or automatic publishing.

    Luma1 source

  5. New
    HiggsfieldAds StudioImages

    Higgsfield Ads Studio turns a saved brand kit into batches of static ads

    Higgsfield announced a static-ad workflow that imports brand and product details from a website or accepts them manually. A saved kit supplies copy and visual direction for up to 20 variations per product. Creatives are downloaded for upload to an ad platform; this is not automatic campaign publishing or video generation. Its October 1 article gives no trustworthy publication hour.

    Higgsfield2 sources

  6. New
    OpenAIChatGPT Images shoppingImages

    ChatGPT adds virtual clothing try-ons and saved shopping finds

    On mobile and web, eligible product listings offer an Images-powered try-on using a reference photo. Users can also save products in Favorites or Library folders. The reference image can be changed or deleted in Personalization settings.

    OpenAI1 source

Wednesday, September 30UTC

  1. New
    HiggsfieldIdeogram 4.5 on HiggsfieldImages

    Ideogram 4.5 editing becomes available on Higgsfield

    The September 30 changelog makes Ideogram 4.5 available in Higgsfield's image selector for reference and text edits, at 1K or 2K. The provider integration has a known date even though Ideogram's own model page is undated.

    Higgsfield2 sources

Days are UTC. Some sources report an exact time and others only a date; both are placed on their UTC day.