RunPod for AI agents

RunPod is a GPU cloud platform for running AI and machine learning workloads, letting teams deploy pods, templates, and serverless endpoints. 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.

Connect RunPod See the 13 actions ↓

Developer toolsAPI keyMCP + RESTrunpod.io

AUDIT LOG · RUNPODPOLICY: acme-support

09:41:07 · claude · u_8f2read

runpod.list_cpu_types

List CPU Types✓ allowed · 212ms

09:41:08 · claude · u_8f2read

runpod.get_pod

Get Pod Details✓ allowed · 164ms

09:41:09 · claude · u_8f2write

runpod.create_secret

Create Secret✓ approved · approved by user

09:41:12 · claude · u_8f2destructive

runpod.delete_template

Delete Template✕ blocked · policy: deny

EVERY RUNPOD CALL, ON THE RECORD

01 · USE CASES

What agents do in RunPod.

Check available GPU types

Get GPU types and pricing before choosing hardware for a new training job.

Save a serverless endpoint

Save a serverless endpoint configuration so a model can be called on demand.

Confirm before deleting a template

Get a template's details first, then delete it only after confirming it's not backing an active pod.

02 · ACTIONS

13 RunPod actions, graded by risk.

Every RunPod action is tagged read, write or destructive, so one policy covers the whole app and new actions inherit the right default.

read

Look things up. Allowed by default.

runpod.get_pod
Get Pod Details
runpod.get_myself
Get authenticated user info
runpod.get_gpu_types
Get GPU Types
runpod.list_cpu_types
List CPU Types

write

Create and change things. Allow, or ask the user first.

runpod.create_secret
Create Secret
runpod.create_cluster
Create RunPod Cluster
runpod.update_registry_auth
Update Registry Auth
runpod.update_user_settings
Update User Settings
runpod.save_endpoint
Save Serverless Endpoint
runpod.save_template
Save Template
runpod.save_registry_auth
Save Container Registry Authentication

destructive

Delete, cancel or archive. Ask first, or deny outright.

runpod.delete_template
Delete Template
runpod.delete_registry_auth
Delete Container Registry Authentication

03 · HOW IT WORKS

RunPod in three steps.

01Your users connect RunPodThey add their RunPod api key on Arc0 Connect, under your brand. It goes straight into the vault.
02You set the rulesReads run, writes like “create Secret” can wait for the user, and “delete Template” can be denied outright.
03Any agent can actYour agent calls RunPod through the Arc0 SDK or MCP, and so can Claude, ChatGPT and Cursor. Every call lands on the audit log.

POLICY.TS

await arc0.policies.set('runpod', {
  read: 'allow',
  write: 'ask',        // create_secret
  destructive: 'deny',  // delete_template
})

# Claude Code: the same connection, one URL
$ claude mcp add --transport http arc0 \
    https://mcp.arc0.ai/u/u_8f2

04 · AUTH AND DATA

How RunPod connects.

Users add their RunPod 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 RunPod 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

05 · WORKS WITH

Use RunPod from any agent.

Claude

ChatGPT

Cursor

Codex

VS Code

OpenAI Agents SDK

Claude Agent SDK

Vercel AI SDK

Mastra

LangGraph

06 · RELATED

More developer tools apps.

Addressfinder17 actions

Addresszen2 actions

Alchemy36 actions

ALL DEVELOPER TOOLS APPS →

07 · FAQ

RunPod and Arc0, answered.

Q01

Can I use RunPod with Claude, ChatGPT or Cursor?

Yes. Connect RunPod 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 RunPod actions you allow.

Q02

How do users connect RunPod?

Users add their RunPod 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.

Q03

Which RunPod actions can my agent take?

13 in total: 4 read, 7 write and 2 destructive, such as “create Secret”. Your policies decide which of them each agent may call.

Q04

Can I stop my agent from deleting things in RunPod?

Yes. Actions like “delete Template” 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.

Q05

Can my own backend call RunPod too?

Yes. The same RunPod connection is available over REST and through the Arc0 proxy, so your product and your agent share one connection per user.

Get started

Plug RunPod into your agent.

Your users connect RunPod once, under your brand. Your agent gets 13 actions behind your policies, with every call on the record.

Connect RunPod Browse all apps →

Free to build · MCP + REST · Audit log on every plan