Installation

Pick the surface that matches how your agent runs. All you need after this is a Floe key (Authentication).

SDKs

TypeScript

npm install floe-agent

import { FloeAgent } from "floe-agent";
const agent = new FloeAgent({ apiKey: process.env.FLOE_API_KEY! });

Python

pip install floe-agentkit-actions

import os
from floe_agentkit_actions import FloeAgent
agent = FloeAgent(api_key=os.environ["FLOE_API_KEY"])

MCP (Claude Code, Cursor, Claude Desktop)

No install — point your MCP client at the hosted server:

claude mcp add --transport http floe https://mcp.floelabs.xyz/mcp \
  --header "Authorization: Bearer $FLOE_API_KEY"

Or run it locally: FLOE_API_KEY=... npx -y @floelabs/mcp-server --stdio. See the MCP Server guide.

REST (any language)

Nothing to install — every capability is one HTTP call with a Bearer header:

curl -X POST https://credit-api.floelabs.xyz/v1/chat/completions \
  -H "Authorization: Bearer $FLOE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"openai/gpt-4o","messages":[{"role":"user","content":"hi"}]}'

CLI

npm install -g @floelabs/cli   # the `floe` platform CLI
npx @floelabs/cli init         # or just run the onboarding once, straight from npm

@floelabs/cli (bin floe) is the full platform CLI — setup, agents, keys, budgets, policies, billing, funds, phone, and metered calls, at parity with the dashboard. The floe-agent SDK package ships its own floe-agent bin: the AgentKit-companion CLI for the agent-runtime SDK. See Floe CLI.

Next

• Authentication — your Floe key

• Add Floe to your existing pipeline — the one-line LLM swap

• Quickstart (5 minutes)