# Video generation references: roles, counts and input modes Source: [Video generation references: roles, counts and input modes]() Documentation index: [llms.txt]() Last updated: 2026-10-28 As reviewed on **28 October 2026**, reference media guide a generation; they do not automatically enter your edit or guarantee identity. Choose the exact model and one compatible input mode, assign each reference a job, and check both **count and aggregate duration**. Seedance 2.0 audio needs an image or video; 2.5 has a separate audio-only mode. Max reference input remains blocked in the inspected ChatCut build. ## What has actually been checked? Official specifications were accessed on **27 October**, with pinned integration and saved route evidence reviewed **28 October 2026**. Local controls do not prove a provider accepted every boundary. The separate local previs example below used one guide and seven images; it is not proof of maximum-count entitlement or current production availability. | Model and public ID | Facts and evidence | | --- | --- | | **Seedance 2.0**
`doubao-seedance-2-0-260128` **Availability:** `available_in_inspected_build`
**Restriction reason:** `pricing_not_wired`
**Endpoint verification:** `transport_inconclusive` Pinned integration and saved dev checks; production unverified. Source links describe native specifications. **Reviewed:** 2026-10-28 | **Native documentation:** 4–15s; 480p/720p/1080p/4K. Frames OR up to 9 images, 3 videos, 3 audios. Ordered shots; timestamp timing is not guaranteed.
Native documentation; not an endpoint test; 2026-10-28. [Native specification source 1 (2026-10-27)](https://docs.byteplus.com/en/docs/ModelArk/create-video-generation-task-api) **ChatCut integration:** 480p/720p/1080p exposed. 4K blocked pending pricing support. Audio enabled internally; no audio-off control.
Pinned integration; production unverified; 2026-10-28. [Native specification source 1 (2026-10-27)](https://docs.byteplus.com/en/docs/ModelArk/create-video-generation-task-api) **Endpoint evidence:** Configured Ark checks timed out; current boundary acceptance, including 4K, remains unverified.
Saved development evidence; not production certification; 2026-10-28. [Native specification source 1 (2026-10-27)](https://docs.byteplus.com/en/docs/ModelArk/create-video-generation-task-api) | | **Seedance 2.5**
`doubao-seedance-2-5-260628` **Availability:** `available_in_inspected_build`
**Restriction reason:** `unknown`
**Endpoint verification:** `transport_inconclusive` Pinned integration and saved dev checks; production unverified. Source links describe native specifications. **Reviewed:** 2026-10-28 | **Native documentation:** 4–30s; 480p/720p/1080p. Frames OR up to 30 images, 10 videos, 10 audios, 50 total. Native audio-only, edit, extend and Draft-to-final.
Native documentation; not an endpoint test; 2026-10-28. [Native specification source 1 (2026-10-27)](https://docs.byteplus.com/en/docs/ModelArk/create-video-generation-task-api) **ChatCut integration:** Generation, reference, edit and extend controls exposed. Draft-to-final and audio-off are not exposed by the inspected tool.
Pinned integration; production unverified; 2026-10-28. [Native specification source 1 (2026-10-27)](https://docs.byteplus.com/en/docs/ModelArk/create-video-generation-task-api) **Endpoint evidence:** Ark acceptance and maximum input boundaries unverified after timeouts. A separate local previs run completed; it does not prove all boundaries.
Saved development evidence; not production certification; 2026-10-28. [Native specification source 1 (2026-10-27)](https://docs.byteplus.com/en/docs/ModelArk/create-video-generation-task-api) | | **MiniMax H3**
`MiniMax-H3` **Availability:** `available_in_inspected_build`
**Restriction reason:** `unknown`
**Endpoint verification:** `validation_only` Pinned integration and saved dev checks; production unverified. Source links describe native specifications. **Reviewed:** 2026-10-28 | **Native documentation:** 4–15s; 768P/2K. First, last-only or both; OR up to 9 images, 3 videos, 3 audios, 12 total. Native audio-only references.
Native documentation; not an endpoint test; 2026-10-28. [Native specification source 1 (2026-10-27)](https://platform.minimax.io/docs/api-reference/video-generation-v2-create.md) **ChatCut integration:** Frame and multimodal-reference modes exposed; do not mix them.
Pinned integration; production unverified; 2026-10-28. [Native specification source 1 (2026-10-27)](https://platform.minimax.io/docs/api-reference/video-generation-v2-create.md) **Endpoint evidence:** Duration rejection enumerated 4–15s. Last-only/audio-only probes did not complete generation; ordinary H3 route success remains unverified.
Saved development evidence; not production certification; 2026-10-28. [Native specification source 1 (2026-10-27)](https://platform.minimax.io/docs/api-reference/video-generation-v2-create.md) | | **MiniMax H3 Max**
`MiniMax-H3-Max` **Availability:** `available_in_inspected_build`
**Restriction reason:** `not_exposed`
**Endpoint verification:** `validation_only` Pinned integration and saved dev checks; production unverified. Source links describe native specifications. **Reviewed:** 2026-10-28 | **Native documentation:** 5–15s; 480P/768P. V2 documentation describes frames and multimodal references; no 2K tier.
Native documentation; not an endpoint test; 2026-10-28. [Native specification source 1 (2026-10-27)](https://platform.minimax.io/docs/api-reference/video-generation-v2-create.md) **ChatCut integration:** Text/frame generation exposed; all Max references blocked. A last frame requires a first frame.
Pinned integration; production unverified; 2026-10-28. [Native specification source 1 (2026-10-27)](https://platform.minimax.io/docs/api-reference/video-generation-v2-create.md) **Endpoint evidence:** Exact endpoint explicitly rejected last-only and audio-only input. Reference validators are not successful in-limit generation proof.
Saved development evidence; not production certification; 2026-10-28. [Native specification source 1 (2026-10-27)](https://platform.minimax.io/docs/api-reference/video-generation-v2-create.md) | ## Choose the reference path - **Agent request:** select **+** for files, or use **Selection Mode** to add `@` references to project assets, clips, time, Viewer regions or Transcript text. Explain what each selected item means. - **Video Gen:** open **Video settings → Input** and fill the visible first-frame, last-frame, reference or source-video slots. Prompt wording alone does not place a file in a frame slot. - **Existing project media:** use compatible assets from **My Assets** rather than uploading duplicates. An eligible project video or public resource can be a reference; it does not have to be a completed generation. A prompt attachment and persistent project media have different lifetimes and roles. See [My Assets](https://pr-692-preview.chatcut.dev/docs/my-assets#what-is-the-difference-between-a-prompt-attachment-and-project-media). A Viewer region selected for the Agent does not by itself establish a provider brush/mask input. ## Seedance: counts, durations and modes For **full 2.0, Fast and Mini**, choose text, first frame, first+last frames, or reference mode. A last frame requires a first; frames cannot mix with reference arrays. The reference budget is **9 images, 3 videos, 3 audios**. Each video/audio is 2–15 seconds and the aggregate is at most **15 seconds per type**. Audio references require an image or video reference. Fast/Mini have the same input rules but different output tiers; see their [family reference](https://pr-692-preview.chatcut.dev/docs/models-seedance-2-fast-mini). For **2.5**, the documented/local caps are **30 images, 10 videos, 10 audios, 50 total**. Each video/audio is 2–30 seconds, with **30 seconds total per type**. Audio-only references are documented separately from 2.0. Frames and reference mode still cannot mix. Edit sources have a separate **4–30-second** requirement; do not apply that minimum to every reference clip. Maximum-boundary supplier acceptance remains unverified after Ark timeouts. Both native specifications list images below **30 MB**, videos up to **200 MB**, audio up to **15 MB**, plus a **64 MB aggregate request** budget. ChatCut’s inspected public-resource path separately caps each image at **5 MiB** and video/audio at **50 MiB**. Satisfy every applicable limit; ten valid-length videos can still exceed an aggregate duration or request-size budget. [Official task API](https://docs.byteplus.com/en/docs/ModelArk/create-video-generation-task-api). ## MiniMax and Omni use different reference contracts **H3** exposes up to 9 images, 3 videos and 3 audio files, **12 total**. Native video/audio are each 2–15 seconds, at most 15 seconds total per type; native images are up to 30 MB, videos 50 MB and audio 15 MB, with a 64 MB request budget. Frame and reference modes are exclusive. H3’s documented last-only/audio-only modes have no completed endpoint generation in this review. **H3 Max** blocks all references in the inspected ChatCut integration. Its configured endpoint explicitly requires first with last, and rejects audio-only references despite broader native documentation. A rejection for ten images or four videos establishes a validator, not successful in-limit generation. See [H3 and Max](https://pr-692-preview.chatcut.dev/docs/models-minimax-h3). **Gemini Omni** is `gemini-omni-1.1-flash-preview`, not Veo or Kling. Its pinned path exposes up to three image references or frames and a separate short source-video path; do not import Seedance counts or audio arrays. See the [Omni reference](https://pr-692-preview.chatcut.dev/docs/models-gemini-omni) for its 10-second source guide and measured encoding tolerance. ## Assign identity, camera and sound separately Use an explicit role statement: “Image 1 controls the product silhouette and color; video 1 controls camera path only; image 2 controls the set and light. Do not copy the guide’s proxy bodies or text.” State what may change. A camera guide, subject sheet and mood board can conflict; adding more files does not resolve that conflict. In a **27 October local Seedance 2.5 previs run**, the submitted request used **one 30-second guide and seven images**, requested 30 seconds, 21:9 and 1080p. The prompt assigned camera/cuts/blocking to the guide, six identity sheets to appearance and the seventh image to the diner. A 30-second export completed with **15/17 checks passing**. Side-view proof failed; speaker assignment and no-music audio remained unverified. This supports one smaller-set workflow, not the maximum reference count or guaranteed identity preservation. 1 #### Prepare a small set Choose one task and exact model. Start with one clear subject reference and, only if needed, one camera guide. Check media type, size and measured duration. 2 #### Assign and inspect Put files in the matching visible slots. Describe what each controls and what must remain fixed; choose frames or references, not both. 3 #### Review the returned result Check the current quote before sending. After completion, compare start/middle/end frames with the intended identity, composition and motion; audition sound separately. Place the accepted clip on the timeline and revise one failed constraint at a time. ## Why did an input disappear or get cropped? Switching models or Generate/Edit/Extend can remove incompatible slots. Recheck every input, resolution, duration and estimate. A fit/crop or output-ratio choice can change composition; inspect the preview and returned frames before assuming a reference was lost. Missing subject detail may also be a generation miss: a completed job does not prove identity or exact camera reproduction. Use [workflow failures](https://pr-692-preview.chatcut.dev/docs/workflow-failures) to separate validation, unavailable routes and output-quality problems. ## Frequently asked questions Must a reference video already be a generated ChatCut video? No. Use an eligible project video or public resource in the supported slot. Media, authorization and moderation rules still apply. Can I mix first/last frames with Seedance references? No. Choose frame mode or multimodal references. In the 2.0 family and 2.5, a last frame also requires a first frame. Why does my 2.0 audio-only request fail? The 2.0 family requires at least one image or video alongside reference audio. Seedance 2.5 has a separate native audio-only mode; current Ark boundary acceptance remains unverified. Does a valid count mean my media are valid? No. Per-file size, media duration, aggregate duration, request size and mode restrictions all apply. The maximum reference boundaries were not completed on the timed-out Ark route. Does an identity sheet guarantee the same face? No. It guides the model. Inspect returned frames against the intended face, wardrobe and composition; one local run’s manual comparison is not a general identity guarantee. **Last verified: 28 October 2026.** Public sources: [Seedance task API](https://docs.byteplus.com/en/docs/ModelArk/create-video-generation-task-api), [MiniMax V2 task API](https://platform.minimax.io/docs/api-reference/video-generation-v2-create.md), accessed 27 October. Pinned tool handling, saved route checks and the separately scoped 27 October local previs report were reviewed 28 October. See [model matrix](https://pr-692-preview.chatcut.dev/docs/model-capability-matrix), [video generation](https://pr-692-preview.chatcut.dev/docs/video-generation-models) and [Credits Policy](https://pr-692-preview.chatcut.dev/docs/credits-policy) before a new request.