Retell AI for AI agents
Retell AI builds and runs voice AI agents for phone and web calls, handling call transcripts, knowledge bases, and conversation flows. 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 Retell AI.
Review a call transcript
Get a completed call's transcript to check what the AI agent and customer discussed.
Add a knowledge source
Add a new source to an agent's knowledge base so it can answer questions on an updated topic.
Confirm before deleting an agent
Check an agent's configuration first, then delete it only after confirming it's no longer in use.
67 Retell AI actions, graded by risk.
Every Retell AI action is tagged read, write or destructive, so one policy covers the whole app and new actions inherit the right default.
read
28Look things up. Allowed by default.
- retell_ai.get_chatGet chat details
- retell_ai.get_agentRetrieve details of a specific agent
- retell_ai.list_agentsList agents
- retell_ai.list_voicesList voices
- retell_ai.get_batch_testGet batch test
- retell_ai.get_chat_agentRetrieve details of a specific chat agent
- retell_ai.get_retell_llmRetrieve details of a specific Retell LLM
- retell_ai.list_all_chatsList all chats
- retell_ai.list_test_runsList test runs
- retell_ai.get_concurrencyGet concurrency
- retell_ai.list_batch_testsList batch tests
- retell_ai.list_chat_agentsList chat agents
- retell_ai.list_retell_llmsList Retell LLMs
- retell_ai.get_agent_versionsGet agent versions
- retell_ai.get_knowledge_baseGet knowledge base
- retell_ai.list_knowledge_basesList knowledge bases
write
29Create and change things. Allow, or ask the user first.
- retell_ai.create_chatCreate a new chat session
- retell_ai.update_callUpdate call
- retell_ai.create_agentCreate Voice AI Agent
- retell_ai.update_agentUpdate agent
- retell_ai.create_batch_testCreate Batch Test
- retell_ai.create_chat_agentCreate a new chat agent
- retell_ai.create_retell_llmCreate Retell LLM Response Engine
- retell_ai.update_chat_agentUpdate chat agent
- retell_ai.update_retell_llmUpdate Retell LLM Response Engine
- retell_ai.add_community_voiceAdd community voice
- retell_ai.update_phone_numberUpdate phone number configuration
- retell_ai.update_chat_metadataUpdate chat metadata
- retell_ai.create_a_new_web_callCreate a new web call
- retell_ai.create_knowledge_baseCreate a new knowledge base
- retell_ai.create_chat_completionCreate chat completion
- retell_ai.create_conversation_flowCreate conversation flow
destructive
10Delete, cancel or archive. Ask first, or deny outright.
- retell_ai.delete_callDelete call
- retell_ai.delete_agentDelete agent
- retell_ai.delete_chat_agentDelete chat agent
- retell_ai.delete_retell_llmDelete Retell LLM
- retell_ai.delete_phone_numberDelete phone number
- retell_ai.delete_knowledge_baseDelete knowledge base
- retell_ai.delete_conversation_flowDelete conversation flow
- retell_ai.delete_test_case_definitionDelete test case definition
- retell_ai.delete_knowledge_base_sourceDelete knowledge base source
- retell_ai.delete_conversation_flow_componentDelete conversation flow component
Retell AI in three steps.
- 01Your users connect Retell AIThey add their Retell AI api key on Arc0 Connect, under your brand. It goes straight into the vault.
- 02You set the rulesReads run, writes like “create a new chat session” can wait for the user, and “delete call” can be denied outright.
- 03Any agent can actYour agent calls Retell AI through the Arc0 SDK or MCP, and so can Claude, ChatGPT and Cursor. Every call lands on the audit log.
await arc0.policies.set('retell_ai', { read: 'allow', write: 'ask', // create_chat destructive: 'deny', // delete_call }) # Claude Code: the same connection, one URL $ claude mcp add --transport http arc0 \ https://mcp.arc0.ai/u/u_8f2
How Retell AI connects.
Users add their Retell AI 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 Retell AI 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 Retell AI from any agent.
Retell AI and Arc0, answered.
Can I use Retell AI with Claude, ChatGPT or Cursor?
Yes. Connect Retell AI 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 Retell AI actions you allow.
How do users connect Retell AI?
Users add their Retell AI 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 Retell AI actions can my agent take?
67 in total: 28 read, 29 write and 10 destructive, such as “create a new chat session”. Your policies decide which of them each agent may call.
Can I stop my agent from deleting things in Retell AI?
Yes. Actions like “delete call” are graded destructive. Set destructive actions to deny, or to ask so the user approves each one, and blocked calls still show up on the audit log.
Can my own backend call Retell AI too?
Yes. The same Retell AI connection is available over REST and through the Arc0 proxy, so your product and your agent share one connection per user.
Plug Retell AI into your agent.
Your users connect Retell AI once, under your brand. Your agent gets 67 actions behind your policies, with every call on the record.
Free to build · MCP + REST · Audit log on every plan