Skip to content

Pricing

Estelle writes the change and checks it against your codebase. Bring your own keys, or use the inference included in your plan.

Save 2 months with yearly billing

All prices in USD. Included usage renews monthly.

Pro

$20/mo

Billed monthly

Codebase memory and tools for individual developers.

Included usage

  • Monthly usage for your individual work
  • Codebase navigation and verification
  • Memory and Atlas
  • Research and reranking
  • Prism reviews and self-correction
  • Silos, agents and monitoring
  • Bring your own model keys
Sign up

Ultra

$100/mo

Billed monthly

Everything in Pro, with more capacity.

Included usage

  • 5× the included usage of Pro
  • Higher memory capacity than Pro for research, reviews and agents
  • Every Pro capability included
Sign up

Max

$200/mo

Billed monthly

Everything in Ultra, for your largest agentic workloads.

Included usage

  • 10× the included usage of Pro
  • Our highest individual memory capacity for your largest agentic workloads
  • Every Ultra capability included
Sign up

Team

$100/mo

Per seat · Billed monthly

All of Estelle, for your whole team.

Included usage

  • 5× the included usage of Pro per seat
  • Every individual capability included
  • Shared usage across your team
  • Per-seat billing
Sign up

Enterprise

Contact us for details

For organizations with specific requirements.

Tell us about your team

  • Your team’s workflows
  • Your deployment requirements
Contact us

Compare plans

Every paid plan includes the whole product. What changes is the included usage, the memory capacity, and whether that balance is shared across seats.

Capabilities included in each Estelle plan
CapabilityProUltraMaxTeamEnterprise
What changes between plans
Seats111Per seatCustom
Included usageRenews monthly. The work it covers depends on the models and tasks you run, so it is not a fixed request count.Included5x Pro10x Pro5x Pro, per seatCustom
Memory capacityHow much of the repository, its history and your team's decisions Estelle keeps indexed at once.Included5x Pro10x Pro5x ProCustom
Shared team balanceEvery seat's allowance pools into one balance.
Products, included in every plan
Estelle AxiomOur coding model.
Estelle PrismReview that reads the real code graph, with the repair attached to the finding.
Estelle SilosA cloud workspace with an editor, a terminal and a desktop beside the run.
Estelle AtlasA readable map of the repository, written from the index.
Estelle MemoryDecisions and corrections carried from one session to the next.
Estelle MonitorProduction issues followed back to the code that caused them.
Ask EstelleAsk about the codebase and get an answer with its sources.
Models and access
Estelle API
Estelle CLI
Bring your own keysYour provider bills that inference directly. Estelle does not charge for it again.
Local modelsPoint the CLI at your own endpoint and run the open-weight model you already host.
Integrations
GitHubConnect repositories, open pull requests, and report checks.
SlackBind a channel and work with Estelle where the team already is.
Linear, Jira and NotionStart work from a ticket and carry its context into the change.

Running a model you host yourself? Install the CLI with npm i -g @fatelabs/estelle, or read the source on GitHub.

Frequently asked
questions

Can I code with Estelle, or does it only connect to other agents?

You can code with Estelle. It has its own model and its own inference, so it writes the change, checks it against your indexed repository, and repairs what does not hold up. You can also connect the agents you already use, and Estelle grounds those the same way.

Do you sell my data?

No. We do not sell your data and we do not train models on your code. Your repository content is used to answer your own questions, for your own account. See the privacy policy for how it is stored and how to delete it.

Can I use my own model keys?

Yes. Bring keys from any provider you already pay for and that provider bills the inference directly, with no second charge from us. You do not have to: every paid plan includes Estelle’s own inference, so your plan alone is enough to build with. Navigation and verification read the index and make no model call at all.

What models do you support?

All the frontier providers, the leading open-weight models, and models you host yourself. For a local model, point the CLI at your own endpoint, including an OpenAI-compatible server such as Open WebUI or Ollama. Install it with npm i -g @fatelabs/estelle, or read the source on GitHub.

Do you deploy or host my software?

No. Estelle proposes a change and a human merges it. We do not deploy your application, host it, or push to your production environment. Where the repair loop is switched on, it opens a reviewable pull request; auto-merge is opt-in, off by default, and gated on the verification passing.

What happens if I exceed my limits?

Nothing breaks and nothing is charged without you choosing it. Add usage from Billing or move to a larger plan, and the added usage goes into the same balance. Navigation and verification keep working, because they read the index rather than calling a model.

How does billing work?

Choose monthly or yearly billing. Yearly shows the full annual charge, and your included usage renews monthly. Team is billed per seat, and every seat’s allowance pools into one shared balance. Usage counting is live, but debiting does not yet cover every operation. Some operations are not yet metered.