Reference images in AI video: dated changes
35 entries on references or consistency from 13 of the 17 lines, 34 of them dated 6 December 2024 to 14 September 2026. The table lists each with its line and version; the chart counts them per line. As of 2026-09-25.
1Reading the notes
Reference notes describe keeping a person, object or place the same across shots by giving the model pictures of it. Runway's Gen-4 References in April 2025 is the earliest note here built around consistent characters and places; Veo 3.1 took up to three images in October 2025, Vidu one to seven, HappyHorse up to nine and Wan 2.7 up to five mixed images and videos. Sora's character references and Kling's elements carry the same idea under other names, and several 2026 notes tie a voice to the reference. Firefly's references work differently: a reference video lends a new generation its composition from July 2025 and its camera motion from December 2025.
| Line | Entries |
|---|---|
| Adobe Firefly Video | 2 |
| Google Veo | 2 |
| Grok Imagine | 1 |
| HappyHorse | 3 |
| Kling AI | 5 |
| Luma Ray | 1 |
| MiniMax Hailuo | 2 |
| PixVerse | 5 |
| Runway | 3 |
| Seedance | 1 |
| Sora | 1 |
| Vidu | 3 |
| Wan | 6 |
Inclusion rule. Lines with at least one entry of this kind. Order. Alphabetical by line.
2Every entry
| Date | Line | Version | What changed |
|---|---|---|---|
| 6 December 2024 | Runway | Act-One | Act-One puts a performance onto characters inside existing videos |
| 30 April 2025 | Runway | Gen-4 | Gen-4 References keeps characters and locations consistent |
| 14 May 2025 | Wan | Wan 2.1 | Wan 2.1 VACE combines local editing, repainting, outpainting, extension and image references |
| 25 June 2025 | Runway | Gen-4 | An update to Gen-4 References improves object consistency |
| 17 July 2025 | Adobe Firefly Video | Firefly Video Model | A reference video can pass its composition to a new generation |
| 5 August 2025 | PixVerse | Not named in the note | Sound effects and Fusion (reference to video) arrive |
| 26 August 2025 | Vidu | Not named in the note | Reference video generation from one to seven images |
| 15 October 2025 | Google Veo | Veo 3.1 | Veo 3.1 preview adds extension, up to three reference images and first-and-last-frame input |
| 21 October 2025 | Vidu | Vidu Q2 | Q2 gains reference-to-video, text-to-video and extension |
| 10 November 2025 | Wan | Wan 2.2 | Wan 2.2 Animate moves a character photo with the performance from a reference video |
| 15 December 2025 | Kling AI | Kling Omni | The Omni video model launches on a new API driven by prompt words |
| 16 December 2025 | Adobe Firefly Video | Firefly Video Model | A start frame can be paired with a reference video whose camera motion the generation recreates |
| 16 December 2025 | Wan | Wan 2.6 | Wan 2.6 reference-to-video keeps a person's look and voice, with several characters at once |
| 18 December 2025 | Luma Ray | Ray3 | Ray3 Modify adds keyframe and character reference controls |
| 22 December 2025 | Kling AI | Not named in the note | Motion Control in the API takes a reference image and a reference video together |
| 13 January 2026 | Google Veo | Veo 3.1 | On Vertex AI, reference-to-video takes 9:16 and 1080p or 4K outputs can be upsampled |
| 23 January 2026 | Kling AI | Not named in the note | Other angles of an element can be filled in from its front view |
| 29 January 2026 | Wan | Wan 2.6 | A Flash tier of Wan 2.6 reference-to-video |
| 12 February 2026 | PixVerse | PixVerse V5.6 | Fusion reference-to-video moves to V5.6 |
| 25 February 2026 | Kling AI | Kling 3.0 | Kling 3.0 and 3.0 Omni reach the API; elements can be built from video |
| 12 March 2026 | Sora | Sora 2 | Character references, clips up to 20 seconds, 1080p on Sora 2 Pro, extensions and batch jobs |
| 23 March 2026 | Kling AI | Not named in the note | Elements made from several images can be tied to a voice |
| 3 April 2026 | Wan | Wan 2.7 | Wan 2.7 reference-to-video mixes up to five image or video references and clones a voice timbre |
| 7 April 2026 | PixVerse | PixVerse C1 | The C1 model covers text, image, transition and reference-to-video |
| 26 April 2026 | HappyHorse | HappyHorse 1.0 | HappyHorse 1.0 reference-to-video takes up to nine reference images |
| 26 April 2026 | HappyHorse | HappyHorse 1.0 | HappyHorse 1.0 video edit changes parts of a clip from instructions and up to five images |
| 21 May 2026 | PixVerse | PixVerse V6 | V6 Fusion makes reference names and types optional |
| 16 June 2026 | HappyHorse | HappyHorse 1.1 | HappyHorse 1.1 reference-to-video keeps subject and scene style steadier |
| 26 July 2026 | PixVerse | PixVerse V6 | V6 reference-to-video takes video references in an omni mode |
| 31 July 2026 | Grok Imagine | Grok Imagine 1.5 | Grok Imagine video 1.5 adds reference-to-video with preset voices and native 1080p |
| 31 July 2026 | MiniMax Hailuo | MiniMax H3 | MiniMax H3 reads text, image, video and audio together as creative context |
| 5 August 2026 | MiniMax Hailuo | Hailuo 3.0 | Runway's developer API adds a model it names MiniMax Hailuo 3.0 |
| 6 August 2026 | Wan | Wan 3.0 | Wan 3.0 folds reference, editing, replication and driving into one model, with clips up to 30 seconds |
| 14 September 2026 | Vidu | Vidu Q3 | Vidu Q3 reference-to-video models for drama and for ads are listed on Model Studio |
| Undated version row | Seedance | Seedance 2.0 | Seedance 2.0: 4 to 15 seconds, up to 4K, multimodal references, editing and extension |
Inclusion rule. All entries tagged with this kind of change, including platform arrivals and undated rows. Order. By date, oldest first; undated rows last.
3One line at a time
- HappyHorse, references
- Kling AI, references
- PixVerse, references
- Runway, references
- Vidu, references
- Wan, references