AI employees in one API call.built without code.live in milliseconds.

One POST. Your new hire gets its own endpoint, memory, tools, and isolated environment. That's it.

Pick a model, write instructions, toggle tools — your new hire is live in minutes. No code needed.

Embed AI agents into your product with a single API call — or build them from the dashboard.

terminal

$ curl -X POST https://api.gopilot.dev/v1/agents \
    -H "X-API-Key: $GOPILOT_KEY" \
    -H "Content-Type: application/json" \
    -d '{
      "name": "Support Bot",
      "model": "anthropic/claude-sonnet-4-20250514",
      "description": "Handles customer support for Acme Corp",
      "llm_keys": [{"provider": "anthropic", "api_key": "sk-ant-..."}],
      "tools": {"allow": ["web_search", "knowledge_base"]},
      "webhook_url": "https://your-app.com/agent-events"
    }'
{"id": "dep_8f894578-...", "object": "agent", "name": "Support Bot",
 "status": "provisioning", "url": "https://openclaw-8f894578.agents.gopilot.dev",
 "model": "anthropic/claude-sonnet-4-20250514", "machine_size": "small",
 "created_at": "2026-03-14T10:30:00Z"}

API Reference

OpenAIAnthropicGoogleMistralAny OpenAI-compatible

Create Agent

Describe what you need and we'll configure everything automatically.

Basics

Model & Key

Channels

Identity

Name your agent and tell it what to do

Name

e.g. Support Bot, Research Assistant

What should this agent do?

e.g. Handle customer support for a SaaS billing platform. It should be friendly, know our pricing tiers, and escalate refund requests to a human.

✨AI-assisted configuration

Machine Size

Resources allocated to the agent's VM

Small1 CU

1 vCPU / 2 GB / 10 GB disk

Medium4 CU

2 vCPU / 4 GB / 10 GB disk

Large8 CU

2 vCPU / 8 GB / 20 GB disk

X-Large16 CU

4 vCPU / 16 GB / 20 GB disk

Browse pre-built agents

terminal

$ curl -X POST https://api.gopilot.dev/v1/agents \
    -H "X-API-Key: $GOPILOT_KEY" \
    -H "Content-Type: application/json" \
    -d '{
      "name": "Support Bot",
      "model": "anthropic/claude-sonnet-4-20250514",
      "description": "Handles customer support for Acme Corp",
      "llm_keys": [{"provider": "anthropic", "api_key": "sk-ant-..."}],
      "tools": {"allow": ["web_search", "knowledge_base"]},
      "webhook_url": "https://your-app.com/agent-events"
    }'
{"id": "dep_8f894578-...", "object": "agent", "name": "Support Bot",
 "status": "provisioning", "url": "https://openclaw-8f894578.agents.gopilot.dev",
 "model": "anthropic/claude-sonnet-4-20250514", "machine_size": "small",
 "created_at": "2026-03-14T10:30:00Z"}

API Reference

OpenAIAnthropicGoogleMistralAny OpenAI-compatible

Build agents with an API call.

One POST request. A live agent with its own endpoint, memory, and tools — running in an isolated env.

terminal

$ curl -X POST https://api.gopilot.dev/v1/agents \
    -H "X-API-Key: $GOPILOT_KEY" \
    -H "Content-Type: application/json" \
    -d '{
      "name": "Support Bot",
      "model": "anthropic/claude-sonnet-4-20250514",
      "description": "Handles customer support for Acme Corp",
      "llm_keys": [{"provider": "anthropic", "api_key": "sk-ant-..."}],
      "tools": {"allow": ["web_search", "knowledge_base"]},
      "webhook_url": "https://your-app.com/agent-events"
    }'
{"id": "dep_8f894578-...", "object": "agent", "name": "Support Bot",
 "status": "provisioning", "url": "https://openclaw-8f894578.agents.gopilot.dev",
 "model": "anthropic/claude-sonnet-4-20250514", "machine_size": "small",
 "created_at": "2026-03-14T10:30:00Z"}

API Reference

OpenAIAnthropicGoogleMistralAny OpenAI-compatible

One call. One agent.

Create a fully functional AI agent with a single API call. It gets its own HTTPS endpoint, its own memory, its own tools. OpenAI-compatible, so your existing code just works.

Isolated by default.

Every agent runs in its own virtual machine. Not a container — a dedicated VM with its own kernel. Your customers' data never touches another tenant. Hardware-level isolation, not just namespaces.

You build. We run.

No Kubernetes. No Docker. No infra to manage. We handle orchestration, networking, TLS, scaling, and monitoring. You focus on your product.

What you can build

E-commerce supportEmbed a support agent in your Shopify app. Each store gets its own agent with its own knowledge base.

Embed a support agent in your Shopify app. Each store gets its own agent with its own knowledge base.

Internal toolsSpin up agents for your ops team. One watches Sentry, one monitors deploys, one handles on-call triage.

Spin up agents for your ops team. One watches Sentry, one monitors deploys, one handles on-call triage.

SaaS featuresGive every customer in your platform their own AI assistant. Multi-tenant isolation included.

Give every customer in your platform their own AI assistant. Multi-tenant isolation included.

Automated researchRun 50 research agents in parallel. Each one searches, reads, extracts, and reports back via webhook.

Run 50 research agents in parallel. Each one searches, reads, extracts, and reports back via webhook.

Conversational appsBuild a chatbot product without building agent infrastructure. We handle the hard parts.

Build a chatbot product without building agent infrastructure. We handle the hard parts.

Email triageSet up an agent that reads your inbox, categorizes messages, drafts responses, and flags what needs your attention.

Set up an agent that reads your inbox, categorizes messages, drafts responses, and flags what needs your attention.

SEO monitoringDeploy an agent that audits your website daily and sends you a report of what to fix, ranked by impact.

Deploy an agent that audits your website daily and sends you a report of what to fix, ranked by impact.

Competitor trackingAn agent that watches competitor websites and alerts you when pricing, features, or messaging changes.

An agent that watches competitor websites and alerts you when pricing, features, or messaging changes.

From zero to production in 4 steps

1Get your API key

export GOPILOT_KEY=gopt_live_k8x...

2Create an agent

One POST request. Define the model, instructions, tools, and channels.

3Integrate

Call your agent's endpoint using any OpenAI-compatible client. Or connect it to Slack, email, or your web app.

python

from openai import OpenAI
client = OpenAI(
    base_url="https://api.gopilot.dev/v1/agents/dep_8f894578",
    api_key="gopt_live_your_key_here",
)
response = client.chat.completions.create(
    model="openclaw",
    messages=[{"role": "user", "content": "What's the status of order #1234?"}],
)
print(response.choices[0].message.content)

4Scale

Create hundreds of agents programmatically. Each one isolated. Each one reachable. Usage-based pricing means you only pay for what you use.

Built for production

/v1/chat/completions

OpenAI-compatible endpoint. Drop-in replacement for any client library.

Webhook callbacks

Get notified when agents complete tasks. HMAC-signed payloads.

Persistent memory

Agents remember context across sessions. Configurable retention.

Multi-model

Switch between OpenAI, Anthropic, Google, Mistral, or self-hosted models per agent.

Custom tools

Give agents access to web search, code execution, file handling, APIs, or build your own.

Channel integrations

Connect agents to Slack, WhatsApp, email, web widgets, or expose the raw API.

SDKs: Python · Node.js · Go · cURL

API Reference

Enterprise-grade isolation

VM-level isolation

Every agent runs in its own Firecracker microVM. Dedicated kernel. No shared runtime.

Zero-knowledge keys

LLM API keys are injected into the VM at boot and stripped from our systems. We never store them.

Encrypted by default

TLS on every endpoint. Encrypted storage. All traffic encrypted in transit.

Tenant separation

Your customers' agents are completely isolated from each other. No cross-tenant data paths.

GDPR compliant

SOC 2 Type II in progress

EU data residency available

Simple, transparent pricing

Start with a free 24-hour Pro trial. Scale when you're ready.

Recommended

Pro

Build and deploy together

$29/month

8 CU — 1 Large (2 vCPU, 8 GB) or 2 Medium (2 vCPU, 4 GB)

8 CU running concurrently

All channels supported

Unlimited workspaces

Bring your own LLM key

24-hour free trial

Team

Build smarter with bigger limits

$79/month

24 CU — 3 Large (2 vCPU, 8 GB) or 1 XL (4 vCPU, 16 GB)

24 CU running concurrently

All channels supported

Unlimited workspaces

Platform-provided LLM key

All plans include: VM isolation, persistent memory, multi-model support, full API access.

Not a developer? No problem.

Pre-configured agents for SEO, competitor monitoring, lead qualification, customer support, and more. Deploy from the dashboard in 60 seconds.

Browse agents

Your first agent is 30 seconds away.

Use the API or the dashboard. Free to start.