Pricing

Priced on what costs us money, not on how many people you employ

Retrievals and staleness refreshes are the two things that cost real money to serve, so they are the two things metered. Seats are free on every tier.

Free

$0/mo

Enough to wire it into a real system of record and prove it works.

1 connected system of record
100 condensed memory objects
500 retrievals / month across app, API and MCP
250 staleness refreshes / month
Entitlement enforced at retrieval time, by role
Full MCP server and REST API access
Self-host it yourself with no limits at all

Start free

most teams start here

Contextely Pro

$49/mo

Production volume, near-real-time freshness, and per-scope entitlement.

20 connected systems of record
10,000 condensed memory objects
25,000 retrievals / month
25,000 staleness refreshes / month
TTLs down to 60 seconds for near-real-time working memory
Per-scope entitlement grants, not just roles
Retrieval audit log: every query, every entitlement refusal
200 members

Start free, upgrade in-app

Self-hosted

Run the whole thing yourself. No plan limits, no metering, no phone-home. You bring a Postgres database and an OpenRouter key; the entitlement and freshness engines are identical to the hosted ones because they are the same code.

Self-hosting guide

Contextely Platform

$499 / month

For putting Contextely inside a product you sell. One API call provisions a client, the people inside it and their keys, and every client is billed to you on one invoice.

250,000 retrievals and 40,000 refreshes a month, pooled across every client
Then $1.20 per 1,000 retrievals and $12.50 per 1,000 refreshes
25 client workspaces, each on the full Pro capability set
A cap per client that you set, so one cannot spend the rest

That is about what 10 Pro subscriptions cost for the same volume. Below that many clients the subscriptions are cheaper, and we would rather you bought those.

How white-labelling works

Questions

Questions people actually ask

What counts as a retrieval?: One call to context_search, whether it came from the web app, the REST API, the MCP server or the CLI. Walking to related memory, listing sources and checking usage do not count. Reading a full record back from its source (context_expand) counts against the refresh allowance rather than the retrieval one, because that is the round-trip it actually costs.
What happens when I hit the limit?: Retrievals return an explicit quota_exceeded error with the reason, rather than silently returning fewer results. An agent can branch on that code and back off or tell its user. Nothing is deleted and nothing degrades quietly.
Why is there a minimum TTL on the free plan?: Because a short TTL is the expensive thing. A one-minute freshness window means Contextely re-reads your system of record and re-condenses constantly, and both of those cost us money per call. Free floors it at an hour; Pro goes down to 60 seconds.
Is entitlement enforcement paywalled?: No. Retrieval-time entitlement is the product and both tiers get it. Free enforces it at role level; Pro adds per-scope grants and the audit log.
What does self-hosting cost?: Nothing, and it has no plan limits. You supply a Postgres database and an OpenRouter key for condensation. The trade is that you run it.
Can I cancel?: Any time, from the billing page, through Stripe's own portal. You drop to Free at the end of the period and keep every memory object you have: nothing is deleted and nothing is pruned. The Free limits then apply to what a query scans, which is the 100 most recently updated, and no new objects are stored until you are back under the limit.