OpenReason shows you what your team spends on AI coding tools, and what that money shipped.

One command on each engineer's machine, and the tools report what they cost. Spend by person, by repo and by pull request, in one place.

Start free — no card See how it works

Works with

Claude Code
Codex
Cursor

Example team

Engineering leadMIMaya Iyer

How it works

Create a team
You get one command.
2
Each engineer runs it once
It configures their coding tools to report to us over OpenTelemetry, and installs a small reporter. Takes a minute.
```
$ curl -fsSL https://get.openreason.app/otel \    | OR_API_KEY=sk-or_... bash  ✓ Setup complete.
```
Example
Data starts arriving
Tokens, model, session length, repo, the commands the agent ran, the prompts.

Your requests still go straight to Anthropic and OpenAI, on your own plans. We are not a proxy.

One command removes all of it.

What each person spent, and what it shipped

One row per engineer. What they spent, what the plan covered, what came out of it.

Example team

Engineering leadMIMaya Iyer

EveryoneLast 30 days

Priya Nandakumar$1,040 · 61%
Tom Okafor$980 · 55%
Lena Brandt$910 · 58%
Dev Raman$870 · 31%
Sofia Mendes$760 · 54%
Arun Prakash$620 · 52%
Mei Tanaka$580 · 56%
Jonas Weber$560 · 50%
Rhea Kapoor$540 · 48%
Kwame Boadi$530 · 44%
Ines Duarte$520 · 39%
Pat Lawson$490 · 42%

ended in a commit or PR
produced nothing

Person	Spent	On plan / billed per request	PRs	Per PR	Shipped	Risk
Priya Nandakumar	$1,040	$950 / $90	14	$74	61%	3
Tom Okafor	$980	$945 / $35	9	$109	55%	1
Lena Brandt	$910	$900 / $10	11	$83	58%	—
Dev Raman	$870	$775 / $95	4	$218	31%	5
Sofia Mendes	$760	$735 / $25	8	$95	54%	2
and 7 more
Team median	$600	$593 / $10	5	$110	51%	1

Priya Nandakumar$1,040$950 on plan · $90 per request14 PRs · $74 each · 61% shipped · 3 risk

Tom Okafor$980$945 on plan · $35 per request9 PRs · $109 each · 55% shipped · 1 risk

Dev Raman$870$775 on plan · $95 per request4 PRs · $218 each · 31% shipped · 5 risk

and 9 more

Team median$6005 PRs · $110 each · 51% shipped · 1 risk

Example team · 12 engineers · last 30 days

Every number sits next to the team median. The same money breaks down by repo, by pull request and by tool.

What your agents did that nobody asked for

176

secrets reached a model last month

Nobody typed them. The agent put them in commands it ran.

Password or secret value148
API key or signing secret24
Database or service URL with a password4
12 people

987tool calls in a repo you don't owndev-raman/ledger-scratch
2times production data was pulled onto a laptopPlus 110 other agent actions.
9MCP servers nobody approved
16prompts were job hunting2.3% of all prompts weren't work.

Example team · 12 engineers · last 30 days

You control what data is collected and how it is stored

Set data collection and retention for your organization

Turn prompt collection on or off, and choose how long prompt text is kept.

Example team

Engineering leadMIMaya Iyer

Privacy

Store prompt and reply text

On: the text of each prompt and reply is kept with its event. Admins can read it, and each person can read their own.

Off: new events are kept without the text. Text stored before you turn it off stays.

Prompt capture is on.

Always kept, whichever you choose

Which agent and model, and when
Tokens, cost and how long each request took
Tool calls, commands and repo names
Risk checks, such as a secret found (its kind, never the value)

Example team · 12 engineers

Start free

OpenReason shows you what your team spends on AI coding tools, and what that money shipped.

Create a team. Run one command. The numbers start arriving.

Start free — no card

OpenReason shows you what your team spends on AI coding tools, and what that money shipped.

How it works

Create a team

Each engineer runs it once

Data starts arriving

What each person spent, and what it shipped

What your agents did that nobody asked for

You control what data is collected and how it is stored

Start free