Unforgetter is opening a small private alpha. There is no price during the alpha and nothing is sold on this site. Inventing three price boxes would look mature and be untrue. This page tells you what alpha access includes, what participation involves, and how pricing will be decided later.
Private alpha
Join the private alpha if you rely on AI agents for ongoing work and want decisions, preferences and current status to carry forward without re-explaining them. One agent is enough; the same memory follows when you switch.
| Private alpha | |
|---|---|
| Who it is for | Anyone who relies on AI agents for work that keeps going and wants a memory that stays current on its own. One agent is enough, and it becomes even more useful when you switch agents or use several |
| Workspaces | One workspace with one owner. Team roles are in build |
| Agents | Compatible with Claude, ChatGPT / Codex, Cursor and Grok Bot through MCP |
| Memory | Versioned Markdown working truth with scopes, provenance and proposals |
| Onboarding | Guided. A fit call, then we connect your agent with you and verify the right workspace comes back |
| Support | Direct contact with the person building it. Expect fast, honest answers and occasional rough edges |
| Data | You choose what is connected. Access is scoped to workspace, project and data class, and what an agent receives carries its provenance |
| Price | No price during the private alpha, and nothing is sold or collected on this site. Public pricing is decided later |
Not finished
Today the connection is made with you: we take the quickstart for your client, add the scoped endpoint and token, and check that the right workspace comes back. A signup flow, OAuth login and a connection assistant are being built.
Membership, invitations and per-member permissions.
Product functions in the alpha foundation, gated on live smoke tests.
Decided after the alpha shows what people value and what it costs to run.
Pricing principles
These are commitments about shape, not numbers. They are here so you can judge the direction before a plan exists.
Adding or switching agent should never be a pricing decision. The value is one memory serving all of them.
Your memory is Markdown. Taking it with you is a product function at every tier, including after you leave.
Limits and caps will be grounded in measured usage before a plan is published.
They are the test of whether the product works. That is why there is no price during the private alpha.
Questions
There is no price during the private alpha. Access is by invitation, nothing is sold or collected on this website, and public pricing is decided later. Participation involves three things: a fit call, a guided connection to your workspace, and feedback from using the product.
No. One agent is enough, because the memory keeps itself current between your sessions. It becomes even more valuable if you use several agents or expect to switch, because they all read the same record.
After the alpha has produced two things: evidence that the product holds up for external workspaces, and measured cost per active workspace. Not before.
As a single-owner workspace, yes. Team roles are in build; you will be told what that means for you before you start.
Say so in the application. Security requirements are read as requirements, not objections.
Every application gets a personal reply. If the fit is wrong, you will be told why.