Try the memory layer on your own coding sessions. It stops at the allowance until you upgrade.
- Tokens ingested
- 10M
- Tokens served
- 3M
- Queries
- 300
- Storage
- 1 GB
- Data uploads
- Not included
Stops at the allowance; upgrade to continue.
Your coding agent runs on the subscription you already have. A Metantel plan covers the memory it reads and writes, metered in tokens, and the same plan works in Flow Console and in Artel.
Checking whether you are signed in…
Sign in to choose or change your plan.
Use the account you sign in to Flow Console or Artel with: the plan belongs to that account.
Signed in as
Try the memory layer on your own coding sessions. It stops at the allowance until you upgrade.
Stops at the allowance; upgrade to continue.
Memory for a working developer, built from your agents' own sessions. Past the allowance you pay as you go.
Beyond the allowance: pay as you go from usage credits.
Shared spaces, data uploads, connected sources and client workflows. Opening once we understand how teams use it.
Beyond the allowance: pay as you go from usage credits.
The memory layer inside your own cloud, with custom workflows and connectors.
On Individual, use past the month's allowance is paid from usage credits you buy in advance; without credits it stops at the allowance, as Free does. The same credits pay for Metantel's models in Flow Console, at each model's price per million tokens. Prices in US dollars.
Buy usage credits| Tokens ingested | $0.033 per million | Your agents' sessions and standard documents, written into memory |
|---|---|---|
| Tokens served | $0.15 per million | What memory hands back to your agents |
| Queries | $0.60 per thousand | Each time an agent asks memory something |
| Storage | $0.20 per GB-month | What your spaces hold, measured nightly |
| Premium research | $2.40 per million served | EDIP's own judged research loop, at least 1,500 tokens a step; off unless you turn it on |
| Premium indexing | $1.60 per million ingested | Cards and full Connectome tagging for documents |
Premium research and premium indexing are never part of an allowance. Your agents browsing the memory catalog themselves are ordinary queries.
AI workers run on your own Claude Code subscription and licence. We charge for the harness and the memory, so your token savings go straight back into your budget.
Artel shows what you have used of each allowance this month, by space and by day, and what use beyond it has come to so far.
Cancel Individual whenever you like: it runs to the end of the month you paid for, then moves to Free. Flow Console and Artel pick up a change as soon as you switch back to them.