AI&Design
All archives
MAY 2024 — NOVEMBER 2026 · KST

Veo

Trace Google’s Veo from its first limited preview to native sound, reference-led generation, and resolution controls. Compare what Google, Runway, and Higgsfield make available.

Posts archived
Attachments examined
Capabilities with evidence

What changed, when

Announcement dates in KST · Select a change or open its example.

Result example attached Announcement or contextual mediaBlank lane · no qualifying announcement in this source

Workflow threads

Read the dated updates and sources
  1. · Google introduces Veo video generation

    Google demonstrates high-definition generation from detailed prompts with support for cinematic styles. Veo was offered to selected creators in a VideoFX private preview through a waitlist.

  2. · Veo 2 begins a wider VideoFX rollout

    Google demonstrates stronger motion and physical realism and control over cinematography, including lens and camera direction. The notice describes a gradual VideoFX waitlist rollout. Maximum model capability and practical product settings differ.

  3. · Veo 3 generates speech and sound with the picture

    Google adds dialogue, ambient sound, and effects to generated videos. Flow launches for cinematic workflows. Google’s I/O release also demonstrates camera controls and reference features of Veo 2. Initial Veo 3 consumer access was limited.

  4. · Google presents Veo 2 camera and reference controls

    Guide Veo 2 using images of a subject, object, scene, or style, and use camera presets to steer the shot. Google presents these demonstrations explicitly under Veo 2 alongside the Veo 3 reveal.

  5. · Gemini animates a photo into a clip with sound

    Upload a still image and direct an eight-second animated result with audio in the Gemini app. Google lists availability for eligible AI Pro and Ultra subscribers where the video feature had rolled out.

  6. · Runway Dev adds Veo 3.1 and Fast

    Generate video from text or an image through Runway’s Dev API using Google Veo 3.1. Runway’s current model and input docs define the route-specific settings and charges.

  7. · Higgsfield documents Veo 3.1 frame and editing workflows

    Higgsfield’s official guide demonstrates generating from start and end frames and preserving existing action when changing part of a shot. Higgsfield displays October 14; its embedded UTC timestamp converts to October 15 KST. The guide shows documented workflows on this surface.

  8. · Veo 3.1 brings sound to framed and extended shots

    Flow’s Ingredients, Frames, and Extend workflows gain generated audio with Veo 3.1. Flow also highlights tools for adding or removing elements. Feature availability on Flow and the Gemini API differs; some Flow controls were experimental.

  9. · Veo 3.1 adds native vertical delivery

    Google improves consistency from reference images and rolls out native 9:16 generation with 1080p or 4K options on supported interfaces. Google describes 4K as an enhancement workflow and specifies which platforms get each output option.

  10. · Higgsfield adds a 4K selection

    Choose 4K in the Veo 3.1 quality settings in Higgsfield Video. The notice covers the product’s 4K setting.

  11. · Veo 3.1 Lite adds a lower-cost generation route

    Google launches its cheaper Veo variant for high-volume work at 720p and 1080p, with text and image inputs and native audio. Lite was a paid-tier Gemini API and Google AI Studio rollout. Supported options differ from other Veo 3.1 tiers.

  12. · Higgsfield adds Veo 3.1 Lite

    Higgsfield’s model picker adds Lite for text- or image-led clips up to eight seconds, with 720p or 1080p and native audio. The release notice includes other models. These settings describe the Veo entry.

  13. · Runway adds negative prompts to Veo routes

    Describe unwanted output in a negative prompt on supported Veo text-to-video and image-to-video requests. The optional field is up to 1,000 characters; Runway applies a default when it is omitted.

  14. · Veo 3.1 Preview exits the Gemini API

    Google shuts down the Veo 3.1, Fast, and Lite preview identifiers in the Gemini API. Google’s notice points developers to Gemini Omni or GA Veo models on Gemini Enterprise Agent Platform. Third-party availability is tracked separately.

Documentation and availability notes
  • Google DeepMind · · The live model page labels demonstrations by version and capability. Each dated event retains the launch source for that capability.
  • Runway Dev API · · Runway documents mode-specific image and frame requirements alongside Veo 3.1 and Fast.
Coverage through November 20, 2026. Exact publisher timestamps are shown in KST; date-only notices retain the publisher's day.Back to capability timeline