Platform launches April 22 — accounts, cross-device sync, Sentinel & teams. Join the waitlist

Simple Pricing

Free for individuals. Sentinel on Vibe, Pro & Team. No surprises.

Free

$0/mo
Get started, no card required
  • 5,000 observations
  • 1 device
  • Cloud sync included
  • Local FTS5 search
  • Secret scrubbing
  • 2 help packs
  • 30-day retention
Get Started

Team

£19/seat/mo
Shared knowledge + Sentinel (min 3 seats)
  • Unlimited observations
  • Unlimited devices
  • Cloud sync
  • Hybrid search (local + cloud)
  • Secret scrubbing
  • All help packs
  • Unlimited retention
  • Sentinel blocking mode
  • Unlimited audits
  • Priority sync
  • Shared team namespace
  • Shared standards library
  • Audit reports
Start Team Trial

Feature comparison

Free Vibe Pro Team
Observations5,00025,000100,000Unlimited
Devices135Unlimited
Agents per deviceUnlimitedUnlimitedUnlimitedUnlimited
Cloud sync
Offline search
Secret scrubbing
Help packs2AllAllAll
Activity heatmapBasicFullFullFull
Retention30 days90 days1 yearUnlimited
Sentinel modeAdvisoryAdvisoryBlocking
Audits/day50200Unlimited
Priority sync
Team namespace
Shared standards
Audit reports

Frequently asked questions

What counts as an observation?

Any piece of knowledge your AI agent captures: a discovery, decision, bugfix, pattern, or change. Each observation is a single structured record with a title, narrative, and metadata. Digests (compacted summaries) also count.

Do early adopters really get free accounts?

Yes. Everyone who joins the waitlist before launch gets a permanent free account with the same limits as the Free tier. No time limit, no catch.

What is the Vibe plan?

The Vibe plan is designed for developers who ship fast and want more than the basics. At $9/mo you get hybrid search, all help packs, Sentinel advisory mode with 50 audits/day, and 90-day retention — the sweet spot between free and pro.

Can I self-host?

The MCP server and local database run entirely on your machine. For sync and team features, you can either use our hosted Candengo Vector service or deploy your own instance with a deployment key.

What happens when I hit my observation limit?

Your agent can still read and search existing observations. New observations will queue locally and sync when you upgrade or when lifecycle compaction frees up space.

What LLM providers does Sentinel support?

Any OpenAI-compatible API: OpenAI, Anthropic, Grok (xAI), Mistral, Ollama (local), vLLM (local), or any custom endpoint. You provide the API key; Sentinel does the rest.

Is my data private?

Your local SQLite database is yours. Secret scrubbing strips API keys and credentials before storage. Sync uses encrypted connections. Team observations respect visibility rules — personal stays personal.

What's the difference between advisory and blocking mode?

Advisory (Vibe and Pro) shows warnings in the Sentinel feed but lets the agent proceed. Blocking (Team) returns exit code 2, which stops the tool call and forces the agent to fix the issue before continuing.

Need more?

Contact us for Enterprise pricing with custom limits, SSO, SLA, and self-hosted deployment.

Contact Us