Ollama for AI agents
Ollama runs large language models locally or in the cloud and exposes them through an OpenAI-compatible API. Developers use it to chat with, generate text from, and manage models without a hosted provider. 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 Ollama.
Generate text with a local model
Send a prompt to a running Ollama model and return the generated completion.
List installed models read-only
Check which models are available locally before routing a request to one of them.
Inspect a model's details
Show a model's parameters and version information before using it in a workflow.
8 Ollama actions, graded by risk.
Every Ollama action is tagged read, write or destructive, so one policy covers the whole app and new actions inherit the right default.
read
2Look things up. Allowed by default.
- ollama.list_modelsList Models
- ollama.showShow Model Information
write
6Create and change things. Allow, or ask the user first.
- ollama.chatChat with Ollama model
- ollama.versionGet Ollama Version
- ollama.generateGenerate Text with Ollama
- ollama.open_ai_completionsOpenAI-Compatible Text Completion
- ollama.open_ai_list_modelsList Models (OpenAI Compatible)
- ollama.open_ai_chat_completionsOpenAI-Compatible Chat Completion
destructive
0Delete, cancel or archive. Ask first, or deny outright.
- No destructive actions.
Ollama in three steps.
- 01Your users connect OllamaThey add their Ollama api key on Arc0 Connect, under your brand. It goes straight into the vault.
- 02You set the rulesReads run, and writes like “chat with Ollama model” can wait for the user to approve.
- 03Any agent can actYour agent calls Ollama through the Arc0 SDK or MCP, and so can Claude, ChatGPT and Cursor. Every call lands on the audit log.
await arc0.policies.set('ollama', { read: 'allow', write: 'ask', // chat destructive: 'deny', }) # Claude Code: the same connection, one URL $ claude mcp add --transport http arc0 \ https://mcp.arc0.ai/u/u_8f2
How Ollama connects.
Users add their Ollama 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 Ollama 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 Ollama from any agent.
ClaudeChatGPTCursor
Codex
VS CodeOpenAI Agents SDKClaude Agent SDK
Vercel AI SDK
Mastra
LangGraph
Ollama and Arc0, answered.
Can I use Ollama with Claude, ChatGPT or Cursor?
Yes. Connect Ollama 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 Ollama actions you allow.
How do users connect Ollama?
Users add their Ollama 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 Ollama actions can my agent take?
8 in total: 2 read, 6 write and 0 destructive, such as “chat with Ollama model”. Your policies decide which of them each agent may call.
Can I make my agent read-only in Ollama?
Yes. Allow read actions and deny writes in the Ollama policy. Your agent can still look things up, and any write it attempts is blocked and logged.
Can my own backend call Ollama too?
Yes. The same Ollama connection is available over REST and through the Arc0 proxy, so your product and your agent share one connection per user.
Plug Ollama into your agent.
Your users connect Ollama once, under your brand. Your agent gets 8 actions behind your policies, with every call on the record.
Free to build · MCP + REST · Audit log on every plan