Ollama Cloud models now run inside Fly.io Sprites via encrypted gateway

Ollama Cloud now works inside Fly.io Sprites, full Linux virtual machines with persistent disks designed for AI agents. The integration, announced October 13, 2026, lets agents in Sprites call Ollama cloud models like kimi-k3, glm-5.3 and deepseek-v4.1-flash through a new Ollama Connector.

How the gateway works

Fly.io encrypts the Ollama API key stored in an organization and never returns it through the API. When code inside a Sprite calls Ollama, the request goes through a Fly.io gateway that checks which Sprite is calling and adds the key to the request. The Sprite only knows a gateway URL and a connection ID, so an agent cannot leak the key.

Access is blocked by default for every Sprite. Administrators choose which Sprites can use the connector by label, by name prefix, or all of them. Calls only work from inside a Sprite, and there is no Authorization header because the gateway adds the key.

Setup and usage

To get started, add an Ollama API key to the Fly.io organization, open Sprites, then Connectors, and add an Ollama connector. Under Access Configuration, select which Sprites can use it and save. A Test reachability button checks that a Sprite can reach Ollama.

From inside a Sprite, a curl command hits the gateway URL with the connection ID filled in, using the standard chat completions endpoint. Any OpenAI SDK works by setting base_url to the gateway and using any value for api_key. Ollama's native /api/* routes also work through the gateway.

Fly.io Sprites come with a skill that tells Claude Code, Cursor, Codex and Gemini CLI about Ollama. If the Sprite has access, the agent can use Ollama models without additional setup.

Example: fuzzing

Fly.io published a guide that uses several models together. A small model like nemotron-3-nano:30b writes malformed inputs, a Sprite runs them against a program, and a larger model like kimi-k3 examines crashes. Switching models only requires changing the model field.

The Ollama Connector quickstart on Fly.io provides more details, along with Ollama Cloud models and pricing pages.