Changelog

You can subscribe to project export completion instead of polling the export GET. Every voiceId field also accepts a catalog displayName (for example Matilda or Charlie) in addition to a vg_voic_… id from GET /v1/resources/tts-voices. Existing clients that send ids keep working.

Signed file download and thumbnail URLs in API responses now use VideoGen hosts. uploadUrl for PUT uploads is unchanged.

Version 2.1 of the TypeScript SDK, Python SDK, CLI, MCP server, and OpenClaw plugin. List VideoGen’s built-in catalog on the Entities API, generate spoken lip-synced clips, theme slideshow-to-video decks, and install the OpenClaw plugin from npm or ClawHub.

Slideshow-to-video no longer requires a theme when you are converting an uploaded PDF or PowerPoint’s original pages.

Slideshow-to-video now requires a slideshow theme so every generated slide shares one design system.

POST /v1/tools/generate-image now accepts optional actor, product, and visual-style entity ids. The model uses them as identity/reference the same way in-app image generation does.

POST /v1/tools/generate-video-clip now accepts spokenDialogue (the exact line the subject should speak as native, lip-synced speech) and voiceDescription (how that voice should sound). The model synthesizes the voice from that text. audioFileIds still accepts a reference recording for lip-sync from that file. You can pass spoken text, reference audio, or both.

POST /v1/tools/generate-video-clip now accepts suppressBackgroundMusic. When true, the generated clip will not include a musical soundtrack. Spoken dialogue and environmental sound are still allowed. Use this when you will add background music separately, for example at the project level.

You can now apply a Ken Burns zoom to every still image in a project with a remix action.

Avatar generation now uses reusable ACTOR entities exclusively.

This release lets MCP hosts identify themselves so connected integrations show up correctly in VideoGen.

This release adds actor-based avatar generation, team file controls, interactive CLI sign-in, and more control over motion graphics. It also expands the agent integration packages and removes several plan restrictions.

This release removes the unsupported workflow-suggestion override from assistant chat creation.

This release fixes MCP OAuth in Cursor so protected tools receive the access token after Sign in with VideoGen.

This release lets storyboard-to-video pick scene lengths automatically, improves the hosted MCP experience in ChatGPT, and updates the in-app URL where you create API keys.

This release adds the Entities API to the docs and SDKs, ships a ChatGPT MCP connector, expands the MCP tool surface, and removes an unsupported workflow method from the public clients.

This release reshapes transcripts around timed words, exposes both the structured transcript and its plain text on file reads, adds caller-provided scene ranges to script-to-video, marks the upload transcript and voiceover scene-range fields as stable, and refines the Assistant API.

This release adds two workflows and a motion-graphic tool, opens up the VideoGen AI assistant as an asynchronous conversational Assistant API, adds timeline interchange for handing projects off to desktop editors, expands export and file controls (including deliver-on-completion destinations), lets you attach transcripts and scene ranges on voiceover runs, exposes structured entity attachments on storyboard scenes, removes the EDIT_WITH_AGENT remix action, and replaces the TypeScript, Python, and CLI clients with rewritten 2.0.0 packages.

This release adds endpoints for verifying your API key and listing past workflow runs and tool executions, fixes project and export ids so responses match the documented vg_proj_... and vg_expo_... format, and ships six new remix actions plus a CLI and MCP server for agents and scripts.

• GET /v1/me: return the account and team behind the API key making the request (apiKeyId, apiKeyNickname, email, displayName, teamId). Takes no parameters. Use it as a connection test to confirm a key is valid before wiring up integrations.
• GET /v1/workflows/runs: list workflow runs started via the API, most recently created first. Cursor-paginated; accepts limit, cursor, and selfOnly (same contract as GET /v1/projects).
• GET /v1/tools/executions: list tool executions started via the API, most recently created first. Cursor-paginated; accepts limit, cursor, and selfOnly.
• GET /v1/resources/languages: list the languages a project can be translated into. Pass a languageCode from the response to the TRANSLATE_PROJECT remix action. Returns the full catalogue in one response (not paginated).
• Six new remix actions: REGENERATE_IMAGES (re-generate every eligible image in a project), UPSCALE_ASSETS (upscale images and videos in place), CHANGE_NARRATOR (swap the voiceover to a different TTS voice), SHUFFLE_STOCK_VISUALS (replace stock footage clips with fresh matches), GENERATE_MUSIC (generate and set a background music track), and TRANSLATE_PROJECT (translate every piece of text and optionally re-narrate voiceovers in a target language). See Remix actions.
• Fix: project and export ids now use the documented opaque format. Responses from GET /v1/projects, workflow runs, remix calls, and export endpoints now return vg_proj_... project ids and vg_expo_... export ids instead of raw UUIDs. Path parameters accept both the new opaque ids and legacy raw UUIDs stored from earlier responses.
• VideoGen CLI: install @videogen/cli (npm install -g @videogen/cli) for a self-contained binary generated from the same OpenAPI spec as the SDKs. See CLI.
• MCP server for AI agents: point Cursor, Claude Desktop, or any MCP client at the VideoGen MCP server to run the full VideoGen API from an agent session (workflows, media tools, files, exports, and remix). Available as a hosted remote server (recommended, no install) or the local @videogen/mcp package. See Use with AI agents.

This release renames the generate-text model field to quality, makes the text-to-speech voice required, adds an image quality tier to the script and voiceover workflows, and introduces three new remix actions. Update your generate-text and text-to-speech calls before upgrading.

• Breaking: generate-text renames model to quality and adds a MAX tier. Replace the model field with quality. Accepted values are now LOW, STANDARD, HIGH, and MAX (previously LOW, STANDARD, HIGH); it still defaults to STANDARD.
• Breaking: text-to-speech now requires voiceId. Pass a voiceId from GET /v1/resources/tts-voices (only voices with supportsDirectToolExecution set to true are accepted). The previous default-voice fallback has been removed, so requests without voiceId now fail.
• Image quality tier on the script and voiceover workflows: POST /v1/workflows/script-to-video and POST /v1/workflows/voiceover-to-video now accept quality (LOW, STANDARD, or HIGH, defaults to STANDARD), the tier used for AI-generated visuals. It only applies when visualStyle.type is AI_IMAGE or ENTITY; STOCK footage is unaffected.
• New remix action RESIZE_PROJECT: change a project’s output aspect ratio and re-flow the video to the new ratio (for example, to a vertical 9:16 social format). See Remix actions.
• New remix action CLEAN_UP_TRANSCRIPT: tighten every transcript in a project by removing filler words and silent pauses, with optional removeFillers, removePauses, fillerWords, and minPauseSeconds controls.
• New remix action CONVERT_IMAGES_TO_VIDEOS: animate every eligible still image into a short AI video clip in place, with optional motionPrompt, muteOutputVideos, and quality controls. It runs asynchronously (one clip per image) and is skipped without error when a project has no eligible images.