Veqen puts every AI provider and agent your company pays for in one record, from your work and coding agents to your product and your teams. It names the person, customer or project behind each charge, tells you, usually within the hour, when a key is used by someone who isn't you, shuts that key off, checks each bill against the provider's own rate card, and books the month for your accountant.

Connect your billing in five minutes, with no code and no card, at any spend size.

Veqen only reads by default and never sees prompts, and it shuts a key off only when you turn that on.

The account is illustrative: when a key spends in a way nobody on the team did, the gap is flagged and labelled, and the key is shut off.

Each provider and agent now shows its own spend, and some cap it, inside its own product. Veqen is the part none of them can do alone, because none of them sees the others.

The same record of your AI bill, read by the people who need it.

One bill per person and team across ChatGPT, Claude, Copilot, Claude Code, Cursor, Codex, Devin and the rest, at the rates you actually pay. Limits hold inside the tools that take them, and you hear as soon as a tool reports someone spending far past their usual.

Cost and margin per customer and feature, and fraud protection for the keys your product runs on: unexplained spend caught, usually within the hour, the key shut off, and the evidence ready for a refund.

Chargeback to cost centers, every invoice checked line by line, overcharges claimed back, and a timeline for every incident your auditor or insurer asks about.

Research labs, internal agents, and document and media pipelines work the same way: anything that calls a model, on any provider or your own GPUs.

Add read-only keys for your providers and your work and coding agents. It takes five minutes and no code.

Every charge is matched to the person, customer or project that made it, key by key and hour by hour, and anything nobody made is flagged with the reason.

Alerts usually arrive within the hour, limits hold inside the tools that take them, a key is shut off in one click, and what you were overcharged becomes a claim.

import veqen from openai import OpenAI

veqen.init() # reads VEQEN_API_KEY; without one, prices locally client = veqen.wrap(OpenAI(), feature="support-bot")

with veqen.tags(customer=account.id): reply = client.responses.create(model="gpt-5-mini", input=question)

Copyimport OpenAI from "openai"; import { wrap } from "veqen";

// Any OpenAI-compatible host is priced as that host const together = wrap(new OpenAI({ baseURL: "https://api.together.xyz/v1" }), { project: "protein-design" });

const reply = await together.chat.completions.create({ model: "Qwen/Qwen3.5-397B-A17B", messages });

Copy# Already emitting GenAI spans (Vercel AI SDK, OpenLLMetry, Logfire)?

Point the exporter at Veqen.

OTEL_EXPORTER_OTLP_TRACES_ENDPOINT=https://veqen-app.com/v1/otlp/v1/traces OTEL_EXPORTER_OTLP_PROTOCOL=http/json OTEL_EXPORTER_OTLP_HEADERS="Authorization=Bearer vq_live_..."

The free audit

Connect read-only access to your providers and your work and coding agents, and within the hour Veqen reports the last 30 days: keys that spent when nobody on your team did, charges above each provider's own rate card, keys owned by people who've left, and seats nobody uses.

Every figure carries the price-book version that priced it, so a monthly total can be traced back to individual requests and tied to an invoice.

ChatGPT, Claude, Copilot, Claude Code, Cursor, Codex, Devin and every other agent in one table per person and team, priced at your contract, with a forecast to month end and the seats nobody uses.

Veqen matches what each provider billed, key by key and hour by hour, with what your code, your gateway and your people actually did. Anything left over is flagged with the reason, so a leak, a mistake and a theft each read as what they are.

shut off 14:02

Shut it off

Turn the key off from the alert, in Slack or the dashboard, with a second person confirming if you like. Veqen keeps the hour-by-hour timeline your provider asks for when you request a refund.

prod-ingest at Anthropic spent $412.10 in 40 minutes. A key nobody used for 30 days was used, on a model it has never run.

Wherever a provider reports both its bill and its usage per key, today OpenAI and Anthropic, each day's charge is checked against what the calls cost at its own rate card, cache, batch and long-context rules included, and overcharges become claims you can file. Coding tools' invoices are checked against your contracts.

With a read-only Stripe key, revenue sits next to AI cost for each customer. Accounts that cost more than they pay are listed first, and the ones that look like abuse are flagged for your fraud tool.

Cost per cost center and team, journal entries for QuickBooks or Xero, and credits from claims counted as they arrive. Veqen for finance and security

Also priced: Embeddings 74Realtime voice 22Image 176Video 44Text to speech 20Speech to text 68Rerank 26Document OCR 16

The dashboard has the full picture. Day to day, most of your team will see Veqen in the tools below, and nearly everything the dashboard does is in the REST API.

An incident arrives with the reason it was flagged and two buttons, which open it ready to shut the key off or to mark it as yours.

Possible stolen key · prod-ingest [ Shut off key ] [ This was us ]

Terminal

Turn your coding tools' exports into a bill per team, on your machine, with nothing uploaded.

npx @veqen/sdk coding-bill
cursor.csv copilot.csv

Pull requests

A cost review on every PR that changes a model, at your real production volume.

gpt-4o → gpt-5-mini monthly impact: −$1,912.40

AI assistants

Ask about spend over MCP from Claude, ChatGPT or Cursor, and set budgets and record bills from the same chat.

"What did each team spend on work and coding agents last month?"

Frequently asked questions

Do I have to change any code?Not for the audit or work and coding agents, which connect with read-only admin keys. To see cost per customer or feature in your own product, add the SDK, which is two lines around the AI client you already use.

Only once you add a separate shut-off credential under Settings › Protection, and Veqen uses it only to turn a key off or back on, when a person asks or a policy you set says so. Where a provider can only delete keys for good, Veqen asks first every time and never deletes by policy. The keys you connect for reading change nothing, and every shut-off is logged.

No. From providers and agents it reads billing, usage and key lists: requests, token counts, models and costs. The SDK sends the same usage fields and the tags you set, and has no field for prompts, responses or code.

On paid plans Veqen reads per-key usage every fifteen minutes from OpenAI, Anthropic, Fireworks AI, fal and Baseten. A key woken after a month unused, a model it has never run, or an hour far above its usual is flagged as soon as the provider reports that hour, which is usually within the hour. Spend that only shows up against your own tracked calls waits two hours for those calls to arrive. At other providers Veqen watches the key list and its totals. Owners, admins and security members get the alert by email, Slack or webhook as soon as it's flagged.

What your AI providers bill, whether or not your code uses the SDK. For each provider Veqen takes what it billed or what Veqen tracked from your calls, whichever is more. It sets your product inference price, and requests themselves are never charged for.

Anyone who used a connected work or coding agent during the billing period. Someone who holds a seat but never used it isn't counted, and shows up under Seats as money you could save.

Yes. Work and coding agents works without the SDK, and product inference works without any agent connected. When you use both, the whole subscription is 15% off.

A claim with the day-by-day evidence the provider asks for: what it billed against its own usage report priced at its rate card, which Veqen checks wherever a provider reports both, today OpenAI and Anthropic. Overcharge checks and claims come with every paid product-inference rung, Veqen takes no share, and every credit you win back is yours, recorded as a provider credit that the month-end close counts.

Billing and per-key usage by the hour, which every protection rule reads: OpenAI, Anthropic, Fireworks AI, fal and Baseten. Billing, or key lists and their totals, for the rules about the keys themselves: OpenRouter, DeepInfra, ElevenLabs, LiteLLM proxy, Vercel AI Gateway, Deepgram, xAI, Google Cloud and AWS Bedrock. Billing alone: Azure. For work and coding agents: Claude Code, Claude, ChatGPT Enterprise, Microsoft 365 Copilot, Gemini in Google Workspace, Codex, Cursor, GitHub Copilot, Windsurf, Devin, Factory, Augment Code, Warp and Ona by API, Kiro, Amazon Q Developer, Amp, Gemini Code Assist, JetBrains AI, Tabnine, GitLab Duo, Sourcegraph Cody and Replit Agent from their exports, and agents such as OpenCode, Aider and Cline from the keys they run on. The SDK prices over 2,000 models, and models on your own GPUs at the price you give them.

Each provider and agent shows its own spend, and some now cap it per project or person, all inside their own product. None of them sees the others, so none gives one bill per person across ChatGPT, Claude Code, Cursor and Devin, and none can tell your traffic from someone else's on the same key. Veqen covers every provider and agent in one place, works key by key and hour by hour, catches spend nobody on your team made while it's still small, shuts off only that key, checks each bill against the provider's own rate card, and books the month for your accountant.

Let your team and your product use all the AI they need, without the bill you didn't see coming

The audit is free at any spend size. Work and coding agents is free up to 3 people, product inference is free up to $1,000 of spend a month, and hourly protection and the month-end close start at $39 a month.