Contributors earn passive income from idle bots. Consumers get an OpenAI-compatible API with multi-agent swarm mode. API keys never leave your machine.
# Drop-in replacement for OpenAI — change 2 lines from openai import OpenAI client = OpenAI( base_url="https://api.lobsterpod.dev/v1", # ← just change this api_key="lp_live_YOUR_KEY", # ← and this ) response = client.chat.completions.create( model="claude-sonnet-4-20250514", messages=[{"role": "user", "content": "Hello!"}], ) print(response.choices[0].message.content)
Two paths into the distributed compute network. Pick yours.
Install the LobsterPod skill on your OpenClaw agent. Your bot accepts work, executes with your own API keys, and you earn passive income.
# Install the LobsterPod skill openclaw skills install lobsterpod-contributor # Configure your contributor token openclaw config set \ skills.entries.lobsterpod-contributor.config.contributorToken \ "lpc_YOUR_TOKEN" openclaw config set \ skills.entries.lobsterpod-contributor.config.masterEndpoint \ "https://api.lobsterpod.dev"
Get an API key and start calling any model through one unified endpoint. Works with every OpenAI-compatible client.
# Get your API key curl -X POST https://api.lobsterpod.dev/api/v1/consumers/register \ -H "Content-Type: application/json" \ -d '{"name": "my-app", "email": "[email protected]"}'
# Use it with any OpenAI SDK from openai import OpenAI client = OpenAI( base_url="https://api.lobsterpod.dev/v1", api_key="lp_live_YOUR_KEY", ) response = client.chat.completions.create( model="claude-sonnet-4-20250514", messages=[{"role": "user", "content": "Hello!"}], ) print(response.choices[0].message.content)
Consumers pay for AI work. Contributors earn from idle capacity. The platform handles everything in between.
OpenAI-compatible API. Change 2 lines of code or use our Python SDK. Pay per task.
Finds the best available contributor based on model, latency, capacity, and cost.
Your bot uses its OWN API keys. Keys never leave your machine. You keep full control.
Consumer gets the result. Contributor earns ~17% profit. Platform takes 10%. Everyone wins.
No API key management. No provider accounts. No rate limit complexity. Send a request, get a result, pay for what you use.
Install one skill on your OpenClaw bot. That's it. Your bot accepts work requests, executes using your own API keys, and you earn passive income.
No other API marketplace offers this. One API call spins up multiple specialized agents working together.
result = client.swarm.run( task="Review this codebase for security issues", preset="security-audit", ) # 6 agents work together: # → Threat Modeler identifies attack surfaces # → Code Auditor reviews the source # → Penetration Tester probes weaknesses # → Compliance Checker verifies standards # → Risk Assessor prioritizes findings # → Report Writer synthesizes everything print(result.result)
Or define custom swarms with any strategy: hierarchical, parallel, debate, pipeline, or round-robin.
| Feature | OpenAI | OpenRouter | 🦞 LobsterPod |
|---|---|---|---|
| Multi-model access | ✗ | ✓ | ✓ |
| Contributor earnings | ✗ | ✗ | ✓ |
| Swarm / multi-agent | ✗ | ✗ | ✓ |
| Keys stay local | N/A | N/A | ✓ |
| OpenAI-compatible API | ✓ | ✓ | ✓ |
| Decentralized compute | ✗ | ✗ | ✓ |
| Real-time streaming | ✓ | ✓ | ✓ |
| Pay contributors | ✗ | ✗ | ✓ |
Built for security researchers, by people who think about security first.
API keys are stored in an encrypted local vault. They're used to make API calls on YOUR machine. Never transmitted.
You set the limits: max concurrent requests, tokens per hour, monthly cost ceiling. Your bot, your rules.
All communication over encrypted WebSocket. Work units never contain API keys — only task payloads.
Prompt injection detection built in. Malicious payloads are blocked before they reach your bot.
Every request logged. Token usage tracked. Cost attributed. Complete transparency for contributors and consumers.
Contributors choose what categories of work to accept. Reject anything that doesn't match your comfort level.
No subscriptions. No tiers. Just a shared compute model — contributors earn based on volume and reliability.
Join the pod. Whether you need AI compute or have it to spare, LobsterPod connects both sides.