Simple Pricing That Scales With You
Transparent, pay-as-you-go pricing across our Unified Intelligence Layer. Universal Knowledge and Memory scale with usage, while LLM Inference is passed through at cost.
Free
For prototypes & evaluation.
$0forever
- 5,000 Knowledge & Memory calls / mo
- Access to Universal Knowledge
- Global Memory Graph
- 1 API key · community support
Starter
For solo builders going live.
$19/ mo
- 50,000 Knowledge & Memory calls / mo
- $4.99 / 1,000 overage
- Temporal analytics
- 1 project · email support
Pro
Scale with zero commitment.
$49/ mo
- 250,000 Knowledge & Memory calls / mo
- $4.99 / 1,000 overage
- MCP / Tavily / OpenAI shims
- Higher rate limits · priority support
Scale
For high-volume products.
$149/ mo
- 1,000,000 Knowledge & Memory calls / mo
- Volume overage pricing
- Dedicated nodes
- SSO · SOC 2
Enterprise
For enterprises & on-prem.
Custom
- Custom volume pricing
- Self-hosting & private deploy
- Custom nodes
- SLA · dedicated support
Billed monthly to your card. Yearly plans get 2 months free.
WATCHMAN — YOUR AGENTS' MEMORY
Local is free. Forever.
Local storage, local search and local team relay cost nothing
| What's counted | Included | After that |
|---|---|---|
| Local memory | Unlimited | Free |
| Local team relay | Unlimited | Free |
| Cloud sync calls | 10,000 / month | $1.00 per 1,000 |
- Running memory on your own machine is free with no cap — storage, ranked search, the recorder and the guards.
- A team syncing peer to peer over the local relay is also free. You only pay when you want the graph synced through the cloud.
- Cloud sync is the paid part because it is the part we run for you.
RATE LIMITS
- Free
- 10 requests / minute · 500 / day
- Paid
- Raised with your plan — see Billing
NO SALES CALL REQUIRED
Every price here is self-serve. Top up, subscribe to a service and see your usage from the billing tab in your dashboard.
GO TO BILLINGAdditional Services Breakdown
UNIFIED LLM ROUTING
Pass-Through at Cost
We don't mark up inference. Whether you route to OpenAI, Anthropic, or DeepSeek, you pay exactly the provider's token cost.
CLOUD AGENTS
$0.000016 / GB-second
Pay only for the exact compute your autonomous sandboxes consume while running tasks. Zero idle server costs.
VOICE & WEBRTC
$0.04 / minute
Ultra-low latency conversational voice powered by our WebRTC endpoints. Billed to the exact second.
COST PER 1,000 CALLS · vs WEB-SEARCH APIs
The lowest flat rate, and the only one with validated, real-time market data.