Veo for AI agents
Veo is Google's model for generating short, high-fidelity videos with sound from a prompt, used by creative teams producing quick video concepts. Connect it once through Arc0, and your agent, or Claude, ChatGPT and Cursor, can use it through one MCP endpoint, limited to what each user approved.
What agents do in Veo.
Generate a video from a prompt
Create a short video with audio from a text description.
Check a generation's status
Look up whether a video generation job has completed, a read-only check.
Download a finished video
Retrieve the finished video file once a generation operation completes successfully.
5 Veo actions, graded by risk.
Every Veo action is tagged read, write or destructive, so one policy covers the whole app and new actions inherit the right default.
read
3Look things up. Allowed by default.
- veo.list_modelsList Models (Gemini API)
- veo.get_videos_operationGet Videos Operation (Veo)
- veo.download_videoDownload Video (Veo)
write
2Create and change things. Allow, or ask the user first.
- veo.wait_for_videoWait For Video (Veo)
- veo.generate_videosGenerate Videos (Veo)
destructive
0Delete, cancel or archive. Ask first, or deny outright.
- No destructive actions.
Veo in three steps.
- 01Your users connect VeoThey add their Veo api key on Arc0 Connect, under your brand. It goes straight into the vault.
- 02You set the rulesReads run, and writes like “wait For Video (Veo)” can wait for the user to approve.
- 03Any agent can actYour agent calls Veo through the Arc0 SDK or MCP, and so can Claude, ChatGPT and Cursor. Every call lands on the audit log.
await arc0.policies.set('veo', { read: 'allow', write: 'ask', // wait_for_video destructive: 'deny', }) # Claude Code: the same connection, one URL $ claude mcp add --transport http arc0 \ https://mcp.arc0.ai/u/u_8f2
How Veo connects.
Users add their Veo api key on Arc0 Connect. It is encrypted in the vault, never shown to the model, and each user can rotate or revoke it at any time.
The same Veo connection serves your agent over MCP and your own backend over REST and the proxy, so a user connects once. How Arc0 handles credentials →
- AUTH
- API key
- CREDENTIALS
- Per-tenant encrypted vault
- MODEL SEES
- Results only, never credentials
- AUDIT LOG
- Every call, on every plan
Use Veo from any agent.
ClaudeChatGPTCursor
Codex
VS CodeOpenAI Agents SDKClaude Agent SDK
Vercel AI SDK
Mastra
LangGraph
Veo and Arc0, answered.
Can I use Veo with Claude, ChatGPT or Cursor?
Yes. Connect Veo to Arc0 once, then add your Arc0 MCP URL to Claude, ChatGPT, Cursor, Claude Code or any other remote-MCP client. Each assistant only gets the Veo actions you allow.
How do users connect Veo?
Users add their Veo api key on Arc0 Connect. It is encrypted in the vault, never shown to the model, and each user can rotate or revoke it at any time.
Which Veo actions can my agent take?
5 in total: 3 read, 2 write and 0 destructive, such as “wait For Video (Veo)”. Your policies decide which of them each agent may call.
Can I make my agent read-only in Veo?
Yes. Allow read actions and deny writes in the Veo policy. Your agent can still look things up, and any write it attempts is blocked and logged.
Can my own backend call Veo too?
Yes. The same Veo connection is available over REST and through the Arc0 proxy, so your product and your agent share one connection per user.
Plug Veo into your agent.
Your users connect Veo once, under your brand. Your agent gets 5 actions behind your policies, with every call on the record.
Free to build · MCP + REST · Audit log on every plan