Deploy nullclaw (OpenClaw compatible)
Fastest, smallest, and fully autonomous AI assistant infrastructure.
nullclaw
Just deployed
/home/nullclaw/
Deploy and Host NullClaw on Railway
NullClaw is an AI agent gateway compatible with the OpenClaw. Full-featured: 22+ providers, 11 channels, hybrid memory, tools, voice. In a 678 KB binary with ~1 MB peak memory usage.
About Hosting NullClaw
NullClaw runs as a persistent gateway that routes incoming messages through your LLM provider of choice. It's a single static binary with no runtime or framework overhead — starts in under 8 ms and runs comfortably on Railway's smallest instances. All you need is a volume for SQLite persistence and an API key from your provider.
Common Use Cases
- Lightweight OpenClaw-compatible agent gateway as a drop-in alternative to heavier runtimes
- Privacy-first AI agent backend for resource-constrained environments
Dependencies for NullClaw Hosting
- An API key from a supported LLM provider (e.g. OpenRouter, Anthropic, OpenAI)
Deployment Dependencies
Implementation Details
NullClaw reads config from ~/.nullclaw/config.json. This template's entrypoint seeds that file from environment variables on first boot. If the file already exists on the volume, it is not overwritten.
| Variable | Required | Default | Description |
|---|---|---|---|
NULLCLAW_API_KEY | ✅ | — | API key for your LLM provider |
NULLCLAW_PROVIDER | ✅ | openrouter | Provider ID (openrouter, anthropic, openai, groq, mistral, deepseek, gemini, ollama) |
NULLCLAW_MODEL | ❌ | anthropic/claude-sonnet-4 | Model to use for inference |
NULLCLAW_TEMPERATURE | ❌ | 0.7 | Inference temperature |
PORT | ✅ | 8080 | Must match Railway's networking config |
NULLCLAW_MEMORY_BACKEND | ❌ | sqlite | Memory backend |
NULLCLAW_MEMORY_AUTO_SAVE | ❌ | true | Auto-save memory |
NULLCLAW_EMBEDDING_PROVIDER | ❌ | openai | Embedding provider for vector memory |
NULLCLAW_VECTOR_WEIGHT | ❌ | 0.7 | Vector search weight |
NULLCLAW_KEYWORD_WEIGHT | ❌ | 0.3 | Keyword search weight |
NULLCLAW_AUTONOMY_LEVEL | ❌ | supervised | Autonomy level (supervised, full) |
NULLCLAW_WORKSPACE_ONLY | ❌ | true | Restrict agent to workspace |
NULLCLAW_MAX_ACTIONS_PER_HOUR | ❌ | 20 | Action rate limit |
NULLCLAW_MAX_COST_PER_DAY_CENTS | ❌ | 500 | Daily cost cap in cents |
NULLCLAW_SANDBOX_BACKEND | ❌ | auto | Sandbox backend |
NULLCLAW_MAX_MEMORY_MB | ❌ | 512 | Max memory for sandboxed processes |
NULLCLAW_MAX_CPU_PERCENT | ❌ | 80 | Max CPU for sandboxed processes |
NULLCLAW_AUDIT_ENABLED | ❌ | true | Enable audit logging |
NULLCLAW_AUDIT_RETENTION_DAYS | ❌ | 90 | Audit log retention in days |
NULLCLAW_TUNNEL_PROVIDER | ❌ | none | Tunnel provider |
NULLCLAW_IDENTITY_FORMAT | ❌ | openclaw | Identity format |
NULLCLAW_SECRETS_ENCRYPT | ❌ | true | Encrypt secrets at rest |
NULLCLAW_HEARTBEAT_ENABLED | ❌ | false | Enable heartbeat |
NULLCLAW_HEARTBEAT_INTERVAL | ❌ | 30 | Heartbeat interval in minutes |
Why Deploy NullClaw on Railway?
Railway is a singular platform to deploy your infrastructure stack. Railway will host your infrastructure so you don't have to deal with configuration, while allowing you to vertically and horizontally scale it.
By deploying NullClaw on Railway, you are one step closer to supporting a complete full-stack application with minimal burden. Host your servers, databases, AI agents, and more on Railway.
Template Content
nullclaw
kadumedim/nullclawNULLCLAW_API_KEY
API key for your LLM provider
