# AI model capability matrix: versions and ChatCut status Source: [AI model capability matrix: versions and ChatCut status]() Documentation index: [llms.txt]() Last updated: 2026-10-28 As of **28 October 2026**, model documentation, ChatCut controls and a working provider route are separate facts. Full Seedance 2.0 documents native 4K, while the inspected ChatCut integration stops at 1080p pending pricing. Kling generation is suspended. The tables below distinguish exposed controls, saved endpoint results and unresolved checks; they do not certify every option in the deployed app. ## How to read this matrix “Exposed” means present in the inspected integration. A successful small development job proves that model and setting, not every reference limit or output tier. “Unverified” means the check did not establish acceptance or rejection. Official documentation was accessed on **27 October 2026** (Omni model card rechecked **28 October**); saved integration evidence was reviewed on **28 October**. Confirm the model, settings and estimate in your current app before submitting. ## Video, image, music and voice models | Model and public ID | Capabilities, integration 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; text, frames or 9 image / 3 video / 3 audio references.
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) [Native specification source 2 (2026-10-27)](https://docs.volcengine.com/docs/ark/seedance-2-0?lang=zh) **ChatCut integration:** 480p/720p/1080p; 4K blocked pending pricing integration. No output-audio toggle.
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) [Native specification source 2 (2026-10-27)](https://docs.volcengine.com/docs/ark/seedance-2-0?lang=zh) **Endpoint evidence:** Ark acceptance unverified: connection timeouts; no successful 4K job.
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) [Native specification source 2 (2026-10-27)](https://docs.volcengine.com/docs/ark/seedance-2-0?lang=zh) | | **Seedance 2.0 Fast**
`doubao-seedance-2-0-fast-260128` **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:** 4–15s; 480p/720p; same 2.0 input modes and reference limits.
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:** Fast selector and 480p/720p exposed in inspected build.
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:** Current configured-route boundary generation unverified after timeouts.
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 2.0 Mini**
`doubao-seedance-2-0-mini-260615` **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:** 4–15s; 480p/720p; same 2.0 input modes and reference limits.
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:** Mini selector and 480p/720p exposed in inspected build.
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:** Current configured-route boundary generation unverified after timeouts.
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; up to 30 image / 10 video / 10 audio references; native edit, extend and Draft.
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) [Native specification source 2 (2026-10-27)](https://docs.byteplus.com/en/docs/ModelArk/seedance-2-5) **ChatCut integration:** Generate/Edit/Extend controls; 30/10/10 local caps. No Draft input or output-audio toggle in 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) [Native specification source 2 (2026-10-27)](https://docs.byteplus.com/en/docs/ModelArk/seedance-2-5) **Endpoint evidence:** Ark boundary and explicit edit/extend semantics unverified after timeouts. Separate local previs run exported 30s.
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) [Native specification source 2 (2026-10-27)](https://docs.byteplus.com/en/docs/ModelArk/seedance-2-5) | | **Gemini Omni 1.1 Flash**
`gemini-omni-1.1-flash-preview` **Availability:** `available_in_inspected_build`
**Restriction reason:** `unknown`
**Endpoint verification:** `not_tested` Pinned integration and saved dev checks; production unverified. Source links describe native specifications. **Reviewed:** 2026-10-28 | **Native documentation:** Distinct Gemini Omni model; Veo specifications do not apply.
Native documentation; not an endpoint test; 2026-10-28. [Native specification source 1 (2026-10-28)](https://docs.cloud.google.com/gemini-enterprise-agent-platform/models/gemini/omni-1-1-flash) **ChatCut integration:** omni alias; 3–10s; 9:16/16:9; 360p/720p/1080p/4k tiers. Saved guide labels 1080p/4K upscaled; first frame or up to 3 images, and short-source editing.
Pinned integration; production unverified; 2026-10-28. [Native specification source 1 (2026-10-28)](https://docs.cloud.google.com/gemini-enterprise-agent-platform/models/gemini/omni-1-1-flash) **Endpoint evidence:** Static contract exposure; no new successful generation in this review.
Saved development evidence; not production certification; 2026-10-28. [Native specification source 1 (2026-10-28)](https://docs.cloud.google.com/gemini-enterprise-agent-platform/models/gemini/omni-1-1-flash) | | **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 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:** Successful last-only/audio-only endpoint generation not established in saved checks.
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. Native docs describe frames and multimodal 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:** Text/frame generation; references blocked in inspected integration.
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:** Configured Max endpoint explicitly rejected last-only and audio-only. In-limit reference completion 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) | | **Kling 3.0**
`kling-v3` **Availability:** `suspended`
**Restriction reason:** `suspended_in_build`
**Endpoint verification:** `validation_only` Pinned integration and saved dev checks; production unverified. Source links describe native specifications. **Reviewed:** 2026-10-28 | **Native documentation:** 3–15s; native 720p/1080p/4K; first frame optionally with last; 1–6 shots; native audio/off control.
Native documentation; not an endpoint test; 2026-10-28. [Native specification source 1 (2026-10-27)](https://kling.ai/document-api/apiReference/model/textToVideo) [Native specification source 2 (2026-10-27)](https://kling.ai/document-api/apiReference/model/imageToVideo) **ChatCut integration:** New generations suspended; absent from manual picker. Showcase describes previous integration.
Pinned integration; production unverified; 2026-10-28. [Native specification source 1 (2026-10-27)](https://kling.ai/document-api/apiReference/model/textToVideo) [Native specification source 2 (2026-10-27)](https://kling.ai/document-api/apiReference/model/imageToVideo) **Endpoint evidence:** Task-list read succeeded; new-generation entitlement and balance unverified.
Saved development evidence; not production certification; 2026-10-28. [Native specification source 1 (2026-10-27)](https://kling.ai/document-api/apiReference/model/textToVideo) [Native specification source 2 (2026-10-27)](https://kling.ai/document-api/apiReference/model/imageToVideo) | | **Nano Banana Pro**
`gemini-3-pro-image` **Availability:** `available_in_inspected_build`
**Restriction reason:** `unknown`
**Endpoint verification:** `completed_generation` Pinned integration and saved dev checks; production unverified. Source links describe native specifications. **Reviewed:** 2026-10-28 | **Native documentation:** 1K/2K/4K tiers and up to 14 references documented.
Native documentation; not an endpoint test; 2026-10-28. [Native specification source 1 (2026-10-27)](https://ai.google.dev/gemini-api/docs/image-generation) **ChatCut integration:** nano-banana maps to Pro; local reference cap 14. Nano Banana 2 is a different model without a pinned selector.
Pinned integration; production unverified; 2026-10-28. [Native specification source 1 (2026-10-27)](https://ai.google.dev/gemini-api/docs/image-generation) **Endpoint evidence:** Small dev output succeeded; high-resolution/full-reference boundaries unverified. Saved adapter replay can select a thought image rather than final output.
Saved development evidence; not production certification; 2026-10-28. [Native specification source 1 (2026-10-27)](https://ai.google.dev/gemini-api/docs/image-generation) | | **GPT Image 2**
`gpt-image-2` **Availability:** `available_in_inspected_build`
**Restriction reason:** `unknown`
**Endpoint verification:** `completed_generation` Pinned integration and saved dev checks; production unverified. Source links describe native specifications. **Reviewed:** 2026-10-28 | **Native documentation:** Image generation/editing; native reference boundary not proved by malformed-input probes.
Native documentation; not an endpoint test; 2026-10-28. [Native specification source 1 (2026-10-27)](https://developers.openai.com/api/docs/models/gpt-image-2) **ChatCut integration:** 10 combined image references; this endpoint rejects transparent backgrounds.
Pinned integration; production unverified; 2026-10-28. [Native specification source 1 (2026-10-27)](https://developers.openai.com/api/docs/models/gpt-image-2) **Endpoint evidence:** Small dev generation succeeded; large-output generation unmeasured.
Saved development evidence; not production certification; 2026-10-28. [Native specification source 1 (2026-10-27)](https://developers.openai.com/api/docs/models/gpt-image-2) | | **GPT Image 2.5 Flare**
`gpt-image-2.5-flare` **Availability:** `available_in_inspected_build`
**Restriction reason:** `unknown`
**Endpoint verification:** `completed_generation` Pinned integration and saved dev checks; production unverified. Source links describe native specifications. **Reviewed:** 2026-10-28 | **Native documentation:** Image generation/editing; size tiers use pixel budgets.
Native documentation; not an endpoint test; 2026-10-28. [Native specification source 1 (2026-10-27)](https://developers.openai.com/api/docs/models/gpt-image-2.5-flare) **ChatCut integration:** 10 combined references; background control exposed for Image 2.5.
Pinned integration; production unverified; 2026-10-28. [Native specification source 1 (2026-10-27)](https://developers.openai.com/api/docs/models/gpt-image-2.5-flare) **Endpoint evidence:** Small dev output succeeded; transparent and large outputs not completed in this audit.
Saved development evidence; not production certification; 2026-10-28. [Native specification source 1 (2026-10-27)](https://developers.openai.com/api/docs/models/gpt-image-2.5-flare) | | **GPT Image 2.5 Sunburst**
`gpt-image-2.5-sunburst` **Availability:** `available_in_inspected_build`
**Restriction reason:** `unknown`
**Endpoint verification:** `completed_generation` Pinned integration and saved dev checks; production unverified. Source links describe native specifications. **Reviewed:** 2026-10-28 | **Native documentation:** Image generation/editing; size tiers use pixel budgets.
Native documentation; not an endpoint test; 2026-10-28. [Native specification source 1 (2026-10-27)](https://developers.openai.com/api/docs/models/gpt-image-2.5-sunburst) **ChatCut integration:** 10 combined references; background control exposed for Image 2.5.
Pinned integration; production unverified; 2026-10-28. [Native specification source 1 (2026-10-27)](https://developers.openai.com/api/docs/models/gpt-image-2.5-sunburst) **Endpoint evidence:** Small dev output succeeded; transparent and large outputs not completed in this audit.
Saved development evidence; not production certification; 2026-10-28. [Native specification source 1 (2026-10-27)](https://developers.openai.com/api/docs/models/gpt-image-2.5-sunburst) | | **Mureka instrumental**
`mureka-9` **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:** Mureka offers generation and separate editing APIs.
Native documentation; not an endpoint test; 2026-10-28. [Native specification source 1 (2026-10-27)](https://platform.mureka.ai/docs/api/operations/post-v1-song-generate.html) **ChatCut integration:** submit\_music creates one new instrumental; 1,024-character prompt cap. No edit/remix/stem/extend controls.
Pinned integration; production unverified; 2026-10-28. [Native specification source 1 (2026-10-27)](https://platform.mureka.ai/docs/api/operations/post-v1-song-generate.html) **Endpoint evidence:** Maximum generated-track duration and stem completion unverified.
Saved development evidence; not production certification; 2026-10-28. [Native specification source 1 (2026-10-27)](https://platform.mureka.ai/docs/api/operations/post-v1-song-generate.html) | | **Mureka song**
`mureka-9.5` **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:** Song generation and separate native APIs.
Native documentation; not an endpoint test; 2026-10-28. [Native specification source 1 (2026-10-27)](https://platform.mureka.ai/docs/api/operations/post-v1-song-generate.html) **ChatCut integration:** submit\_music creates one new song; 1,024-character prompt / 5,000-character lyric caps.
Pinned integration; production unverified; 2026-10-28. [Native specification source 1 (2026-10-27)](https://platform.mureka.ai/docs/api/operations/post-v1-song-generate.html) **Endpoint evidence:** 9.5 remix explicitly denied on tested route; no established eight-minute song ceiling.
Saved development evidence; not production certification; 2026-10-28. [Native specification source 1 (2026-10-27)](https://platform.mureka.ai/docs/api/operations/post-v1-song-generate.html) | | **ElevenLabs voice and sound**
`eleven_v3` **Availability:** `available_in_inspected_build`
**Restriction reason:** `unknown`
**Endpoint verification:** `not_tested` Pinned integration and saved dev checks; production unverified. Source links describe native specifications. **Reviewed:** 2026-10-28 | **Native documentation:** Model-specific TTS APIs; separate speech-to-speech/dialogue routes. Provider sound validator lists 0.5–30s.
Native documentation; not an endpoint test; 2026-10-28. [Native specification source 1 (2026-10-27)](https://elevenlabs.io/docs/overview/models) **ChatCut integration:** Tagged official voices use eleven\_v3 via TTS; sound effects capped at 0.5–22s. No v4 dialogue or voice-changer control via submit\_voice.
Pinned integration; production unverified; 2026-10-28. [Native specification source 1 (2026-10-27)](https://elevenlabs.io/docs/overview/models) **Endpoint evidence:** Short upstream speech-to-speech succeeded separately; 30s sound effect not generated.
Saved development evidence; not production certification; 2026-10-28. [Native specification source 1 (2026-10-27)](https://elevenlabs.io/docs/overview/models) | ## Does Seedance 2.0 generate native 4K? **Yes for full Seedance 2.0 in the checked official API.** Its BytePlus public ID is `dreamina-seedance-2-0-260128`; the corresponding Ark model is `doubao-seedance-2-0-260128`. ChatCut’s inspected pricing integration supports 480p/720p/1080p and blocks 4K. Configured Ark probes timed out, so 4K acceptance there remains unverified. No current ChatCut 4K price was verified. [Official API](https://docs.byteplus.com/en/docs/ModelArk/create-video-generation-task-api), [Ark tutorial](https://docs.volcengine.com/docs/ark/seedance-2-0?lang=zh). Native generation creates the source frames. Upscaling resizes an existing clip; export renders the timeline at delivery settings. A 4K export alone does not establish native 4K generation. See the [Seedance 2.0 answer](https://pr-692-preview.chatcut.dev/models/seedance-2.0). ## How do full 2.0, Fast, Mini and 2.5 differ? Full 2.0, Fast and Mini document 4–15 seconds. Fast and Mini both offer 480p **and** 720p; neither is the 2.5 Draft API. Full 2.0 adds higher documented tiers, with the ChatCut 4K restriction above. Seedance 2.5 documents 4–30 seconds at up to 1080p. All use text, frame or reference modes, with different budgets. There is no matched four-model quality, speed or current-cost benchmark in this review. See [Fast and Mini limits](https://pr-692-preview.chatcut.dev/docs/models-seedance-2-fast-mini). ## How do Seedance 2.5 references and modes work? The 50-file maximum is **30 images + 10 videos + 10 audio files**, not 50 of any type. Audio-only references are documented natively. Frame mode cannot mix with reference mode; a last frame requires a first frame. Media-duration budgets apply as well as file counts. Native edit/extend selectors and Draft-to-final exist, but the inspected ChatCut tool has Generate/Edit/Extend modes and **no Draft input**. Its edit/extend adapter uses prompt classification rather than forwarding the native explicit task selector; endpoint semantics remain unverified after timeouts. [Official API](https://docs.byteplus.com/en/docs/ModelArk/create-video-generation-task-api). 2.5 supports integer-second timestamp prompting; 2.0 supports ordered shots without a timestamp guarantee. A separate local previs run exported 30 seconds, with 15 of 17 acceptance checks passed; side-view proof failed and speaker/no-music audio remained unresolved. That bounded result does not verify every reference boundary or deployed route. See [Seedance 2.5](https://pr-692-preview.chatcut.dev/models/seedance-2.5). ## Does a last frame always require a first frame? The requirement depends on the model and route. **Seedance frame mode and Kling 3.0 require a first frame when a last frame is supplied.** Native MiniMax H3 documents last-only and audio-only references, but successful endpoint generation in those modes was not established here. The configured **H3 Max** endpoint explicitly rejected both last-only and audio-only input. H3 and H3 Max must remain separate rows. [MiniMax API](https://platform.minimax.io/docs/api-reference/video-generation-v2-create.md). ## Which should I choose: Seedance or Kling? For a job inside the inspected ChatCut integration, choose a model offered in the app: new Kling jobs are suspended. Native Kling 3.0 documents 3–15 seconds, 720p/1080p/4K, 1–6 shots and an audio control; these do not make it callable in ChatCut. A first frame can optionally be paired with a last frame. No matched quality winner or current matched-price comparison was established. The [Kling page](https://pr-692-preview.chatcut.dev/models/kling-3.0) keeps its examples labelled as the previous integration. ## What changes between Seedream versions? Seedream is **not selectable in the inspected ChatCut image tool**. Ark probes timed out; lack of a selector does not mean the native models lack these capabilities. Keep these documented native specifications separate from availability: | Model and public ID | Capabilities, integration and evidence | | --- | --- | | **Seedream 4.0**
`seedream-4-0-250828` **Availability:** `not_offered_yet`
**Restriction reason:** `not_exposed`
**Endpoint verification:** `transport_inconclusive` Pinned integration and saved dev checks; production unverified. Source links describe native specifications. **Reviewed:** 2026-10-28 | **Native documentation:** Up to 14 image references; references + sequence outputs ≤15; 1K/2K/4K.
Native documentation; not an endpoint test; 2026-10-28. [Native specification source 1 (2026-10-27)](https://docs.byteplus.com/en/docs/ModelArk/image-generation-api) **ChatCut integration:** No Seedream selector in inspected image tool.
Pinned integration; production unverified; 2026-10-28. [Native specification source 1 (2026-10-27)](https://docs.byteplus.com/en/docs/ModelArk/image-generation-api) **Endpoint evidence:** Ark acceptance unverified after timeouts.
Saved development evidence; not production certification; 2026-10-28. [Native specification source 1 (2026-10-27)](https://docs.byteplus.com/en/docs/ModelArk/image-generation-api) | | **Seedream 4.5**
`seedream-4-5-251128` **Availability:** `not_offered_yet`
**Restriction reason:** `not_exposed`
**Endpoint verification:** `transport_inconclusive` Pinned integration and saved dev checks; production unverified. Source links describe native specifications. **Reviewed:** 2026-10-28 | **Native documentation:** Up to 14 image references; references + sequence outputs ≤15; 2K/4K.
Native documentation; not an endpoint test; 2026-10-28. [Native specification source 1 (2026-10-27)](https://docs.byteplus.com/en/docs/ModelArk/image-generation-api) **ChatCut integration:** No Seedream selector in inspected image tool.
Pinned integration; production unverified; 2026-10-28. [Native specification source 1 (2026-10-27)](https://docs.byteplus.com/en/docs/ModelArk/image-generation-api) **Endpoint evidence:** Ark acceptance unverified after timeouts.
Saved development evidence; not production certification; 2026-10-28. [Native specification source 1 (2026-10-27)](https://docs.byteplus.com/en/docs/ModelArk/image-generation-api) | | **Seedream 5.0 Pro**
`dola-seedream-5-0-pro-260628` **Availability:** `not_offered_yet`
**Restriction reason:** `not_exposed`
**Endpoint verification:** `transport_inconclusive` Pinned integration and saved dev checks; production unverified. Source links describe native specifications. **Reviewed:** 2026-10-28 | **Native documentation:** Up to 10 references → one image at 1K/1.5K/2K. Layer mode: one input → one base plus up to 16 alpha layers. Ark ID: doubao-seedream-5-0-pro-260628.
Native documentation; not an endpoint test; 2026-10-28. [Native specification source 1 (2026-10-27)](https://docs.byteplus.com/en/docs/ModelArk/image-generation-api) [Native specification source 2 (2026-10-27)](https://docs.byteplus.com/en/docs/ModelArk/seedream-5-0-pro) **ChatCut integration:** No Seedream selector or verified layer-delivery workflow in inspected tool.
Pinned integration; production unverified; 2026-10-28. [Native specification source 1 (2026-10-27)](https://docs.byteplus.com/en/docs/ModelArk/image-generation-api) [Native specification source 2 (2026-10-27)](https://docs.byteplus.com/en/docs/ModelArk/seedream-5-0-pro) **Endpoint evidence:** Ark acceptance unverified after timeouts.
Saved development evidence; not production certification; 2026-10-28. [Native specification source 1 (2026-10-27)](https://docs.byteplus.com/en/docs/ModelArk/image-generation-api) [Native specification source 2 (2026-10-27)](https://docs.byteplus.com/en/docs/ModelArk/seedream-5-0-pro) | 5.0 Pro layer decomposition can return one base plus up to 16 editable alpha layers. The generation tier table does not describe every layer’s dimensions. No ChatCut layer-delivery workflow was verified. Sources: [image API](https://docs.byteplus.com/en/docs/ModelArk/image-generation-api), [5.0 Pro guide](https://docs.byteplus.com/en/docs/ModelArk/seedream-5-0-pro), accessed 27 October 2026; reviewed 28 October. ## Is Gemini Omni the same as Veo? **No.** ChatCut’s `omni` alias identifies `gemini-omni-1.1-flash-preview`. The inspected tool has no Veo adapter or selector. Current Google Cloud Veo specifications, retired Gemini API preview endpoints and a vendor wrapper are separate contracts. | Model and public ID | Capabilities, integration and evidence | | --- | --- | | **Veo 3.1 (Cloud)**
`veo-3.1-generate-001` **Availability:** `not_offered_yet`
**Restriction reason:** `not_exposed`
**Endpoint verification:** `not_tested` Pinned integration and saved dev checks; production unverified. Source links describe native specifications. **Reviewed:** 2026-10-28 | **Native documentation:** 4/6/8s; 720p/1080p/4K; 16:9/9:16; native sound. Text, first, first+last, 1–3 subject images or video extension are separate modes. Subject references require 8s; no audio-reference upload.
Native documentation; not an endpoint test; 2026-10-28. [Native specification source 1 (2026-10-27)](https://docs.cloud.google.com/gemini-enterprise-agent-platform/models/veo/3-1-generate) **ChatCut integration:** No Veo adapter or selector in inspected ChatCut tool; omni is a different model.
Pinned integration; production unverified; 2026-10-28. [Native specification source 1 (2026-10-27)](https://docs.cloud.google.com/gemini-enterprise-agent-platform/models/veo/3-1-generate) **Endpoint evidence:** ChatCut generation access unverified; no supplied Veo E2E.
Saved development evidence; not production certification; 2026-10-28. [Native specification source 1 (2026-10-27)](https://docs.cloud.google.com/gemini-enterprise-agent-platform/models/veo/3-1-generate) | | **Veo 3.1 Fast (Cloud)**
`veo-3.1-fast-generate-001` **Availability:** `not_offered_yet`
**Restriction reason:** `not_exposed`
**Endpoint verification:** `not_tested` Pinned integration and saved dev checks; production unverified. Source links describe native specifications. **Reviewed:** 2026-10-28 | **Native documentation:** 4/6/8s; native sound; model card lists 720p/1080p and reference/extension controls. A 4K pricing entry conflicts with the card.
Native documentation; not an endpoint test; 2026-10-28. [Native specification source 1 (2026-10-27)](https://docs.cloud.google.com/gemini-enterprise-agent-platform/models/veo/3-1-generate) **ChatCut integration:** No Veo adapter or selector in inspected ChatCut tool.
Pinned integration; production unverified; 2026-10-28. [Native specification source 1 (2026-10-27)](https://docs.cloud.google.com/gemini-enterprise-agent-platform/models/veo/3-1-generate) **Endpoint evidence:** 4K needs endpoint confirmation; ChatCut access unverified.
Saved development evidence; not production certification; 2026-10-28. [Native specification source 1 (2026-10-27)](https://docs.cloud.google.com/gemini-enterprise-agent-platform/models/veo/3-1-generate) | | **Veo 3.1 Lite (Cloud preview)**
`veo-3.1-lite-generate-001` **Availability:** `not_offered_yet`
**Restriction reason:** `not_exposed`
**Endpoint verification:** `not_tested` Pinned integration and saved dev checks; production unverified. Source links describe native specifications. **Reviewed:** 2026-10-28 | **Native documentation:** 4/6/8s; 720p/1080p; native sound; no asset references. Image-input, first/last and extension descriptions conflict in checked docs.
Native documentation; not an endpoint test; 2026-10-28. [Native specification source 1 (2026-10-27)](https://docs.cloud.google.com/gemini-enterprise-agent-platform/models/veo/3-1-generate) **ChatCut integration:** No Veo adapter or selector in inspected ChatCut tool.
Pinned integration; production unverified; 2026-10-28. [Native specification source 1 (2026-10-27)](https://docs.cloud.google.com/gemini-enterprise-agent-platform/models/veo/3-1-generate) **Endpoint evidence:** Contradictory native modes and ChatCut access remain unresolved.
Saved development evidence; not production certification; 2026-10-28. [Native specification source 1 (2026-10-27)](https://docs.cloud.google.com/gemini-enterprise-agent-platform/models/veo/3-1-generate) | The Cloud extension guide describes a 1–30s source plus 7 seconds, up to 37s total. Do not apply the old Gemini API extension limit or a wrapper’s controls to this route. The Gemini API Veo 3.1 preview shutdown notice dated 22 October 2026 does not shut down the Cloud GA IDs above. No supplied test establishes ChatCut Veo generation access. Sources: [Cloud model card](https://docs.cloud.google.com/gemini-enterprise-agent-platform/models/veo/3-1-generate), [Cloud extension guide](https://docs.cloud.google.com/gemini-enterprise-agent-platform/models/video/extend-a-veo-video), [Gemini API notice](https://ai.google.dev/gemini-api/docs/veo?hl=en); accessed 27 October, reviewed 28 October 2026. ## Choose settings and review the result 1 #### Choose a task and model Open Video Gen and check the models your current build actually offers. For a short revision, compare supported inputs before selecting the model. 2 #### Use one compatible input mode Start with text, a first frame, first plus last, or the selected model's reference mode. Check file counts and aggregate media duration. 3 #### Review the quote before submitting Choose duration, ratio and resolution, then inspect the current estimate. Wait for completed output and review motion, identity, cuts and audio before editing or exporting. Use [generation references](https://pr-692-preview.chatcut.dev/docs/generation-references) for attachments and [workflow failures](https://pr-692-preview.chatcut.dev/docs/workflow-failures) for recovery. Current tariffs belong to [Credits Policy](https://pr-692-preview.chatcut.dev/docs/credits-policy). Model names alone do not set the price: resolution, duration, video inputs and account cohort matter, and Agent usage is separate. This matrix supplies no current-price quote. ## Model references available in this build - [Seedance 2.0 Fast and Mini: inputs and limits](https://pr-692-preview.chatcut.dev/docs/models-seedance-2-fast-mini) - [GPT Image 2, Flare and Sunburst: controls and route checks](https://pr-692-preview.chatcut.dev/docs/models-gpt-image) - [Nano Banana Pro and Nano Banana 2: identities and limits](https://pr-692-preview.chatcut.dev/docs/models-nano-banana) - [MiniMax H3 vs H3 Max: frames, references and route limits](https://pr-692-preview.chatcut.dev/docs/models-minimax-h3) - [Gemini Omni 1.1 Flash: model identity, inputs and output tiers](https://pr-692-preview.chatcut.dev/docs/models-gemini-omni) The [video generation guide](https://pr-692-preview.chatcut.dev/docs/video-generation-models), [Seedance prompt examples](https://pr-692-preview.chatcut.dev/blog/seedance-2-0-prompts-examples) and model pages above provide the next steps. ## Frequently asked questions Can I generate native 4K with Seedance 2.0 in ChatCut? Full Seedance 2.0 documents native 4K. The inspected ChatCut integration blocks 4K pending pricing support and exposes up to 1080p. The configured Ark endpoint timed out, so its 4K acceptance is unverified. Is Seedance Mini only 720p? No. Mini and Fast both document 480p and 720p for 4–15 seconds. They are distinct from full 2.0 and from Seedance 2.5 Draft. Does a last frame always need a first frame? Seedance and Kling 3.0 require a first frame with a last frame. Native H3 documents last-only, with endpoint success unresolved here. The configured H3 Max endpoint explicitly rejected last-only. Does Seedream 5.0 Pro offer the same 4K output as 4.5? No. The checked native API lists 1K/1.5K/2K for 5.0 Pro and 2K/4K for 4.5. Neither is selectable in the inspected ChatCut image integration. Is 4K export the same as native 4K generation? No. Native generation creates source frames at the model’s output tier. Upscaling resizes an existing clip, and export renders the timeline at delivery settings. A 4K export alone does not establish native 4K source generation. **Last fact review: 28 October 2026.** This page uses dated official documentation, the inspected integration contract and saved development checks. It is not a live certification of every production account or endpoint.