> ## Documentation Index
> Fetch the complete documentation index at: https://docs.vantage.sh/llms.txt
> Use this file to discover all available pages before exploring further.

> ## Agent Instructions
> If the person you are helping needs a new Vantage account, read https://docs.vantage.sh/auth.md before you register one.

# Vantage Tokenomics

> Measure and optimize token spend by attributing AI usage to identity, purpose, and outcome.

> [!NOTE]
>
> Vantage Tokenomics is in Private Preview for select Vantage design partners and will open in Beta to existing customers soon. To request access, contact [support@vantage.sh](mailto:support@vantage.sh) or your Vantage account team. Not a Vantage customer? Contact [sales@vantage.sh](mailto:sales@vantage.sh) or [book a demo](https://www.vantage.sh/contact/sales).

Vantage Tokenomics is a set of tools to measure and optimize token spend, track how efficiently each person uses tokens, and show the value of completed work like GitHub pull requests and Linear issues. Tokenomics automatically attributes estimated AI cost and usage to **identity** (who made the request), **purpose** (what it was for), and **outcome** (what it produced), so you can see who is driving spend and what value it delivers.

Tokenomics works by ingesting OpenTelemetry (OTel) logs from AI IDEs and AI gateways, with no applications installed on developer laptops. Once logs begin syncing, three new areas of the Vantage console are available:

> [card] [Roster](#roster)
>
> Everyone generating token spend, with the tasks they completed and what each cost.

> [card] [Classifier](#classifier)
>
> Your organization's tagging strategy, applied to every request as it happens.

> [card] [Tokenomics Report](#tokenomics-report)
>
> Custom reports on cost, usage, and efficiency to find workflows worth optimizing.

To connect your tools, go to [Get Started](#get-started). For answers about data retention, privacy, and more, see the [Frequently Asked Questions](#frequently-asked-questions). To learn why we built Tokenomics, read the [Introducing Vantage Tokenomics](https://www.vantage.sh/blog/vantage-tokenomics) announcement.

## How Vantage Tokenomics Works

1. **Your AI tools send OpenTelemetry logs to Vantage.** Coding tools and AI gateways export usage over OTel with an ingest token from Vantage. Provider admin APIs supplement this when a tool's native export is not available.
2. **The Classifier tags each request.** Hooks in your coding tools or gateway fetch your organization's tagging skill from Vantage and apply it as work happens, so developers never tag by hand.
3. **Usage appears in the Roster and Tokenomics Reports within minutes.** Vantage estimates the gross cost of each request and calculates efficiency metrics across users, models, and tags.

Vantage installs hooks, skills, and configuration so your coding tools can send OTel data. It does not install applications or run scripts on developer laptops, and it does not intercept network traffic. See [Does Vantage install anything on developer laptops?](https://www.vantage.sh/blog/vantage-tokenomics#faq-laptops)

### Supported Tools and Gateways

Any AI tool that supports OpenTelemetry is compatible with Vantage Tokenomics, so support grows as the industry adopts open standards. Most common tools support OpenTelemetry today:

| Coding tools | AI gateways |
| - | - |
| Cursor | OpenRouter |
| Claude Code | LiteLLM |
| Codex | Bifrost |

If you use a tool that isn't listed, submit a feature request through your Vantage account team or by emailing [support@vantage.sh](mailto:support@vantage.sh).

> [!TIP]
>
> LiteLLM and Bifrost also have [LLM Enrichment](https://docs.vantage.sh/llm_enrichment.md) integrations, which use request logs to allocate billed provider costs in Cost Reports. See [LiteLLM](https://docs.vantage.sh/litellm_enrichment.md) and [Bifrost](https://docs.vantage.sh/bifrost_enrichment.md). To learn how the two features differ, see [How does this relate to Cost Reports and LLM Enrichment?](https://www.vantage.sh/blog/vantage-tokenomics#faq-cost-reports)

## Get Started

Users with the **Organization Owner** or **Integration Owner** role can turn on Tokenomics for an account. See [Role-Based Access Control](https://docs.vantage.sh/rbac.md) for details.

**1. Turn on Tokenomics**

In the Vantage console, open **Tokenomics** and turn it on for your account.

**2. Connect your sources**

Connect your coding tools, your organization-wide AI gateway, or a mix of both, depending on your coverage. Vantage provides onboarding instructions for each source. See [Set Up with an AI Gateway](#set-up-with-an-ai-gateway) and [Set Up Coding Tools Without a Gateway](#set-up-coding-tools-without-a-gateway).

**3. Open the Roster**

Once logs begin to populate, the Roster is your default Tokenomics page. From there, open Tokenomics Reports to build custom views of who is spending, what they are spending on, and the outcomes.

Tokenomics collects data from the time a source is connected. It cannot backfill earlier usage. See [Can I backfill historical data?](https://www.vantage.sh/blog/vantage-tokenomics#faq-backfill)

### Set Up with an AI Gateway

An administrator can set up a gateway once for the whole organization, with nothing installed on developer laptops.

1. Add the gateway from Tokenomics onboarding or the Vantage **Integrations** page, and create an OpenTelemetry ingest token.
2. Add the ingest token to the gateway's OpenTelemetry configuration.
3. Configure the hook that imports your organization's tagging skill. Do this once per gateway. Skill URLs are static, so you don't need to change the configuration when you update your tagging strategy.

For full details, see [How do I set up Tokenomics if my organization uses an AI gateway?](https://www.vantage.sh/blog/vantage-tokenomics#faq-gateway-setup)

### Set Up Coding Tools Without a Gateway

Coding tools connect through the Vantage Tokenomics plugin in the [finops-as-code](https://github.com/vantage-sh/finops-as-code) GitHub marketplace.

* **Enterprise subscriptions:** An administrator installs the plugin once for the organization in Claude Code, Codex, or Cursor. In Cursor, mark the plugin **Required** for the team.
* **Team and individual subscriptions:** Vantage MCP users can call `setup_vantage_tokenomics` to have their agent install and configure the plugin, and `check_vantage_tokenomics` to verify setup at any time. You can also share a setup prompt for each developer to run in their coding agent. See [Vantage MCP](https://docs.vantage.sh/vantage_mcp.md).

For install commands and the setup prompt, see [How do I set up Tokenomics if my organization does not use an AI gateway?](https://www.vantage.sh/blog/vantage-tokenomics#faq-coding-tool-setup)

## Roster

The Roster is the homepage of Tokenomics. It lists everyone in your organization generating token spend over the trailing 30 days, sorted by estimated gross spend, alongside tokens consumed and the GitHub pull requests and Linear issues they completed. You can also sort by the number of pull requests and issues completed. Select a name to see the tasks that person completed and the cost of each.

![The Roster, listing developers by estimated gross spend with the pull requests and issues they completed](https://assets.vantage.sh/blog/vantage-tokenomics/roster.png)

To add someone who isn't on the Roster yet, invite them from the Roster to set up the OpenTelemetry stream and the Classifier.

## Classifier

The Classifier applies your organization's tagging strategy to each request as it happens, including requests that arrive through a gateway, so developers never tag by hand. Classifier tags supplement tags already on your data and do not override them by default.

![The Classifier page, where administrators define the organization's tagging strategy](https://assets.vantage.sh/blog/vantage-tokenomics/classifier.png)

By default, the Classifier applies these tags:

* **Identity:** The user the spend belongs to.
* **Purpose:** What the tokens are used for, such as development, research, design, or bug fixes.
* **Outcome:** The result of the spend, including pull request ID, branch, repository, and ticket ID from systems like Jira or Linear.

You can edit the guidance for the default tags, but you can't delete them. Administrators can add up to 10 more tag keys, such as Project, Business Unit, or Team, and describe how to set them in plain language. Because tagging happens per request, a session that mixes different kinds of work is split across the tags that apply to each request.

Manage your tagging strategy on the **Classifier** page in the Vantage console. Changes apply to future requests and do not reprocess previously tagged costs. See [Where do I manage the Classifier, and how do I make changes?](https://www.vantage.sh/blog/vantage-tokenomics#faq-classifier)

## Tokenomics Report

Tokenomics Reports show estimated cost, usage, and efficiency within minutes, attributed to the user, purpose, and outcome behind them. Chart tokens consumed, estimated gross cost, cost per 1M tokens, cache hit rate, or input-to-output ratio. Group or filter by model, provider, source, token type, or any tag, including the Classifier's default tags and the tags your organization adds.

![A Tokenomics Report charting cost per 1M tokens by user, with efficiency metrics for each developer](https://assets.vantage.sh/blog/vantage-tokenomics/report.png)

Select a user to open their developer page, with their usage and efficiency metrics, unique sessions, and unique outcomes.

## Frequently Asked Questions

The [Introducing Vantage Tokenomics](https://www.vantage.sh/blog/vantage-tokenomics) announcement answers common questions:

* [How much does this cost?](https://www.vantage.sh/blog/vantage-tokenomics#faq-pricing)
* [What tools and gateways are supported?](https://www.vantage.sh/blog/vantage-tokenomics#faq-supported-tools)
* [Does Vantage install anything on developer laptops?](https://www.vantage.sh/blog/vantage-tokenomics#faq-laptops)
* [What can I group and filter by?](https://www.vantage.sh/blog/vantage-tokenomics#faq-group-filter)
* [How are efficiency metrics calculated?](https://www.vantage.sh/blog/vantage-tokenomics#faq-efficiency-metrics)
* [How are estimated costs calculated?](https://www.vantage.sh/blog/vantage-tokenomics#faq-estimated-costs)
* [How fresh is the data, and how long is it retained?](https://www.vantage.sh/blog/vantage-tokenomics#faq-data-retention)
* [Can I backfill historical data?](https://www.vantage.sh/blog/vantage-tokenomics#faq-backfill)
* [Does Vantage collect prompts and completions?](https://www.vantage.sh/blog/vantage-tokenomics#faq-prompts)
* [How does this relate to Cost Reports and LLM Enrichment?](https://www.vantage.sh/blog/vantage-tokenomics#faq-cost-reports)


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.