AI&Design

Capability timeline

Vendor-reported claims · May 18, 2025 to Nov 18, 2026

Midjourney V8.2 · Reference editsSee evidence ↗

Choose dates

Announcement dates

Bars show dated records. Click to center or drag to select. Use the handles to resize or move.

One changeZoom into nearby datesRelease examples

Select a date or a cluster to zoom.

  • Reference consistency
  • Partial edits
  • Resolution
  • Image to video
  • SVG output
  • Drawing interchange

Capability records

Browse dated announcements, model names, and official sources.

Reference consistency · 24 records
  • Higgsfield ·

    A studio that turns a script into a short series: it splits the text into episodes, builds the cast, locations, and props, and generates the episodes with the same characters throughout.

    Official source Open on timeline

  • Higgsfield · Gemini Omni Flash 1.2 ·

    It works from text, images, and video; references keep characters and objects consistent, per Google, and a reference video’s audio is ignored.

    Official source Open on timeline View archive entry

  • ChatGPT image · ChatGPT Images 2.6 ·

    Our new image generation model creates images faster and is better at preserving details across edits, like a character’s outfit when you change the background.

    Official source Open on timeline View archive entry Read the article

  • Higgsfield · Nano Banana 2.1 ·

    References: up to 14 images, with consistency for up to 4 characters and up to 10 objects.

    Official source Open on timeline View archive entry

  • Higgsfield · Nano Banana 2.1 ·

    Compared with Nano Banana 2, it improves visual quality, prompt adherence, character consistency across edits, and text rendering, per Google.

    Official source Open on timeline

  • Higgsfield · Grok Imagine Video 1.5 ·

    References: up to 7 images to keep a character or style consistent across generations.

    Official source Open on timeline

  • Higgsfield · Grok Imagine Video 1.5 ·

    Video extension: continue a clip from its last frame, keeping the character, lighting, and motion consistent.

    Official source Open on timeline View archive entry

  • Higgsfield · GPT Image 2.5 Flare; GPT Image 2.5 Sunburst ·

    Both: sharper detail, better subject preservation from references, targeted edits that hold up across iterations, and stronger style holding.

    Official source Open on timeline View archive entry

  • Higgsfield · MiniMax H3 Max ·

    Image-to-video: animates a source image, keeping its composition, identity, and aspect ratio, per MiniMax.

    Official source Open on timeline View archive entry

  • Higgsfield · Wan 3.0; Wan 3.0 Prime ·

    Consistency: faces, clothing, and details hold across frames, even with several characters.

    Official source Open on timeline

  • Grok · Grok Imagine Image 2.0 ·

    One world from a handful of prompts: a character, her locations, and the props she carries — generated separately, holding one style from image to image.

    Official source Open on timeline View archive entry

  • Grok · Grok Imagine Video 1.5 ·

    Pass in a character image and a voice reference, and both hold — the same face and the same voice in every scene.

    Official source Open on timeline View archive entry

  • Grok · Grok Imagine Video 1.5 ·

    Each reference image locks one thing in place — a face, a product, a location.

    Official source Open on timeline View archive entry

  • Grok · Grok Imagine Video 1.5 ·

    Keep a character and swap the scene, keep the scene and swap the character, or hold both and change only the action.

    Official source Open on timeline

  • Higgsfield · Gemini Omni Flash ·

    Consistent characters: face, outfit, and voice hold across edits, with no recasting.

    Official source Open on timeline

  • Higgsfield ·

    Fights, chases, and explosions are generated with real weight and inertia; face, body, and clothing stay consistent across the video; sound effects, speech, and background music are generated with the video.

    Official source Open on timeline

  • Higgsfield ·

    A 15-image set of the same character in different poses, outfits, and locations, generated in one run.

    Official source Open on timeline

  • Higgsfield · Seedream 5.0 Lite ·

    Consistency: a character’s face, clothing, and details hold better across images.

    Official source Open on timeline

  • Higgsfield · Omni 3.0 ·

    Kling Omni: text, images, video, and audio in one model; generates dialogue, sound effects, and lip-synced speech with the video, keeps characters consistent from references, and handles multi-character scenes.

    Official source Open on timeline View archive entry

  • Runway · Gen-4 References ·

    An updated version of Gen-4 References that significantly improves object consistency and prompt adherence. Available for all users.

    Official source Open on timeline View archive entry Read the article

  • Midjourney · Midjourney V7 ·

    Omni-reference (--oref) lets you put consistent characters and objects into scenes

    Official source Open on timeline View archive entry Read the article

  • Higgsfield · Veo 3.1 · Date unconfirmed

    Upload reference images or videos to lock in a specific style, character, or aesthetic for unlimited generations.

    Official source Open on timeline

  • Higgsfield · Date unconfirmed

    The AI Video Generator lets you replace faces inside cinematic scenes, edit every image with surgical precision, and maintain perfect character and lighting consistency from the first shot to the last.

    Official source Open on timeline

  • Higgsfield · SoulID · Date unconfirmed

    It connects directly with other Higgsfield features like SoulID, which ensures consistent identity across projects, and Sora 2, which adds cinematic realism and motion control.

    Official source Open on timeline

Partial image and video edits · 10 records
Output resolution · 61 records
Image-to-video generation · 27 records
SVG generation · 5 records
Drawing interchange · 6 records
  • Autodesk Forma ·

    Import a DXF into the floor plan and trace over it to build your plan.

    Official source Open on timeline

  • Autodesk Forma ·

    We know some of our customers still prefer the reliable workflows that AutoCAD offer, so now you can take a floor plan design from Forma Building Design into AutoCAD as a DWG file, so you can continue working on your model instead of redrawing it from scratch.

    Official source Open on timeline Read the article

  • Autodesk Forma ·

    You control which buildings, components, labels, and units go into each export, so you can tailor the file to the job: a fuller export to continue technical design, or a lighter one for polished floor plan drawings.

    Official source Open on timeline

  • Autodesk Forma ·

    Totals by type feed into the area metrics panel, and the DXF export carries parking on its own layer.

    Official source Open on timeline

  • Autodesk Forma ·

    Furniture is also included in the Revit export as families or as blocks in DWGs of course.

    Official source Open on timeline

  • Autodesk Forma ·

    Design them in floorplan view, and they appear in the 3D scene, in sections and, of course, in the Revit export.

    Official source Open on timeline

Release examples · 1 records