Pricing
One meter: fields. A field is a store, a team, a project, a client: one container of records, opened by one token.
From DIN 1.0.39 the field count is enforced: din field new refuses a field past
your plan's number, and says which plan, which number and how to move up. Fields you already have
keep working. The ops and client-org figures below are enforced the same way, from your signed licence
(an ops or a client org past the plan is refused); the node figures are shown, not enforced.
You are not billed for volume, for seats or for inference. You buy the number of fields you need kept apart. Free imports and grades your history (up to the Reader level, in a 40,000-character window), places it and recalls it; every paid tier adds live capture, every reading level and the whole-chat read, and the rest of the capability table below, row by row. Self-hosted first: hosting is optional and never forced, and hosted carries a store cap per org: 10 GB on Starter, 25 GB on Freelancer, 50 GB on Team, 100 GB on Small Business.
Grading runs on inference you already have, never ours: your assistant CLI (Claude Code or Codex), your own API key (Anthropic, OpenAI, or any OpenAI-compatible endpoint) or a local model (Ollama, vLLM, llama.cpp). DIN never pays for or sees those calls, and it forecasts the cost before a run. The Codex, key and local paths are built and tested on stand-ins; none of them has been validated against a real vendor yet. So there is no usage line here and no surprise at the end of the month.
It runs on your box. This is the licence.
Set up with you. An isolated tenant on DIN's hosted service, with a store cap per org: 10 GB on Starter, 25 GB on Freelancer, 50 GB on Team, 100 GB on Small Business. Nothing is ever deleted or trimmed automatically.
| Tier | Fields | Ops layers | Orgs | Self-host | |
|---|---|---|---|---|---|
| Free | 1 | 0 | none | $0per month | Download |
| One field, no signup, no card. History import graded up to the Reader level (a 40,000-character window), placement, recall, the scoped brief and the local console; live capture is paid. Grading runs on your own inference: your assistant CLI, your own API key or a local model. Up to 3 people and 1 AI agent in that field. | |||||
| Starter | 6 | 2 | 1 | $29per month | Start |
| Freelancer | 15 | 5 | 1 | $99per month | Start |
| Team | 30 | 10 | 1 | $199per month | Start |
| Small Business | 90 | 30 | 1 | $299per month | Start |
| Agency | 90 | 30 | 5 | $399per month | Contact usSet up with you. |
| Set up with you. The same capacity as Small Business (90 fields, 30 ops), split across up to five client orgs, one per client, five walls, managed from one install with din org. Each org's fields have their own stores, tokens and consoles; a sixth client org is refused until one is archived. Hosted is set up with you too: an isolated tenant on DIN's hosted service, with a store cap of 50 GB per org. | |||||
| Enterprise | 180+ | any | 10+ | quoted | Talk to us |
| 180 fields and up, more client orgs, self-hosted or hosted. Hosted: 100 GB per org. Enterprise is a conversation, not a checkout: ask, and a person answers with a figure for what you need. | |||||
| Tier | Fields | Ops layers | Orgs | Hosted | |
|---|---|---|---|---|---|
| Free | 1 | 0 | none | not sold hosted | Download |
| One field, no signup, no card. History import graded up to the Reader level (a 40,000-character window), placement, recall, the scoped brief and the local console; live capture is paid. Grading runs on your own inference: your assistant CLI, your own API key or a local model. Up to 3 people and 1 AI agent in that field. | |||||
| Starter | 6 | 2 | 1 | $50per month | Contact usSet up with you. |
| Freelancer | 15 | 5 | 1 | $149per month | Contact usSet up with you. |
| Team | 30 | 10 | 1 | $249per month | Contact usSet up with you. |
| Small Business | 90 | 30 | 1 | $349per month | Contact usSet up with you. |
| Agency | 90 | 30 | 5 | $599per month | Contact usSet up with you. |
| Set up with you. The same capacity as Small Business (90 fields, 30 ops), split across up to five client orgs, one per client, five walls, managed from one install with din org. Each org's fields have their own stores, tokens and consoles; a sixth client org is refused until one is archived. Hosted is set up with you too: an isolated tenant on DIN's hosted service, with a store cap of 50 GB per org. | |||||
| Enterprise | 180+ | any | 10+ | quoted | Talk to us |
| 180 fields and up, more client orgs, self-hosted or hosted. Hosted: 100 GB per org. Enterprise is a conversation, not a checkout: ask, and a person answers with a figure for what you need. | |||||
Already a customer? Sign in to your account for your licence key, the download and billing.
Free, paid and hosted, side by side
| Feature | Free | Paid, self-hosted | Hosted |
|---|---|---|---|
| Can I get it today? | Yes: a download, no signup, no card | Starter to Small Business are on sale; for the rest, contact us | Contact us: set up with you, no card |
| The program (binary) | Free | The same program, plus a licence key | We run it for you |
| The source code | Not published yet | Not published yet | Not published yet |
| Where your records live | Your machine | Your machine | An isolated tenant on DIN's hosted service |
| Fields | 1 | 6 to 90 by tier, more quoted | By tier, as self-hosted |
| People and AI agents on one field | Up to 3 people and 1 AI agent | Up to 25 nodes (people, agents, machines) | As self-hosted |
| Grading, placement and recall | Included | Included | Included |
| Graded history import | Included | Included | Included |
| Live capture from your agent (Claude Code, Cursor, OpenCode, Codex), and the grading of those live sessions | Not included | Included | Included |
| Reading depth | Up to Reader: Scanner and Reader, in the fixed 40,000-character window | All five levels, Scanner to Scholar, depending on your model, and the whole-chat read | As self-hosted |
| Trust marks and model provenance on every record | Included | Included | Included |
| The trust-weighted blast radius | Not included | Included | Included |
| The dependency watch (scheduled library checks, upgrade briefs, the suspicious-package guard, the org dependency inventory) | Not included | Included | Included |
| Graded recall | Not included | Included | Included |
| The org roll-up | Not included | Included | Included |
| The org security roll-up | Not included | Included | Included |
| Webhooks | Not included | Included | Included |
| Team review (more than three people stamping a field) | Not included | Included | Included |
| The succession packet (one person's graded work, assembled for a handoff) | Not included | Included | Included |
Enterprise
Enterprise (180 fields and up) is a conversation, not a checkout. The field maths continues past Agency and the shape does not change: more fields, more client orgs, self-hosted or hosted. There is no listed price and nothing to join: ask, and a person answers with a figure for what you need.
The rules, stated simply
din org new). Two or more countries or legal entities is a second command,
and that is a conversation rather than a row.In every tier, including Free
- The record and recall path: placement, recall, the scoped brief (
din brief), and this console. - Graded history import: bring in the history you already have, from any source DIN reads, graded up to the Reader level in the fixed 40,000-character window (below).
- Trust marks on every record: inferred, logged, receipted, never collapsed into one badge.
- Model provenance on every record, and the model-drift view.
- Grading on your own inference: your assistant CLI (Claude Code or Codex), your own API key (Anthropic, OpenAI, or any OpenAI-compatible endpoint) or a local model (Ollama, vLLM, llama.cpp). No model calls on our account and no metered inference line; DIN forecasts the cost first. The Codex, key and local paths are built and tested on stand-ins; none of them has been validated against a real vendor yet.
- Self-hosting. A paid licence is verified when you run
din activate: one call to our licensing service, or none on a machine with no network:din activate --filewith a saved reply, which lasts 14 days too, so an offline install needs a fresh reply carried to it every 14 days. The programs then check its signature on your machine. A paid install checks in withapi.dinctrl.com, with a random install id, when you rundin activate --renewand at most once a day when the field or the console starts (or withdin autostart); each signed check-in lasts 14 days, and an install that goes 14 days without one reads as Free until its next check-in. A free install never checks in. A key stays valid for up to 72 hours past each billing date. - On a self-hosted install, your records are never deleted on a downgrade or a lapse: they are files on your machine.
In every paid plan
- Live capture: new work captured as it happens, from Claude Code, Cursor, OpenCode or Codex,
several agents on one field, and the why-layer's grading of those live sessions (
din grade on). History import stays free. - Every reading level, Scanner to Scholar, depending on your model, and the whole-chat read: the record read whole, in one call, when it fits the model (below).
- The trust-weighted blast radius.
- Graded recall.
- The org roll-up.
- The org security roll-up.
- Webhooks.
- Team review (more than three people stamping a field).
- The succession packet (one person's graded work, assembled for a handoff).
- Dependency watch, the delivery: scheduled checks of the libraries you use for vulnerabilities, deprecation
and abandonment, upgrade briefs with blast radius, the suspicious-package guard, and the org dependency
inventory (
din deps refresh,guard,brief,timer, the catalog fetch and check, and--api). The delivery is paid, not the data: the library inventory (din deps sync,din deps inventory) and the local risk report (din dep-audit, anddin dep-audit --enrichby hand) are free on every plan. The delivery is included in every paid plan, not sold separately. The scheduled check is off until you switch it on; by default it answers from DIN's signed library catalog, so no package name leaves your machine, and asking the public registries directly is a separate opt-in that sends package names and versions only. The catalog covers eight ecosystems: Rust (cargo), npm, Python (PyPI), Go, Ruby (RubyGems), PHP (Packagist), .NET (NuGet) and Java (Maven). It holds the registry's complete name list for cargo, PyPI, RubyGems and Packagist; for npm, Go, NuGet and Maven it holds the packages indexed so far, and a package it does not hold is reported as not in the catalog, not as safe.
Reading levels: how deeply a grade reads
A grade is as deep as its read. DIN names five reading levels, by what they read, not by whose model:
- Scanner: what was plainly said: statements, obvious names, listed items. Breadth, not depth.
- Reader: the substance, one section at a time. It can lose the threads that run across sections.
- Analyst: the whole thread: decisions, outcomes, most specifics.
- Archivist: the complete, citeable record: every fact quote-backed, across every field.
- Scholar: everything the Archivist keeps, plus the deepest reasoning across it: contradictions, cross-references, what changed and why.
Free tops out at Reader: it grades in a fixed 40,000-character window, in pieces. Paid unlocks all five, depending on your model. Analyst and above need the record read whole, in one call; a chat read in pieces stays below Analyst. Which of the top three a run reaches is your model's own level.
Prices in USD, per month. Taxes may apply at checkout.