Specivo
AI usage & spend

See where your AI budget actually goes.

Specivo shows how much each person on your team spends on AI, and on which project, feature and issue. Start at the company total and open it down to a single agent session.

  • Collected automatically from Claude Code, Codex, Gemini CLI and OpenCode
  • Every session tied to the issue it worked on
  • Runs on your own server; only usage numbers leave the laptop
See the live demo Self-host in 15 minutes

Demo data: ACME Waste & Recycling (illustrative)Spend by person and feature

The problem

The invoice says how much. Not on what.

Your teams use Claude Code, Codex, Cursor and other AI tools every day, on personal plans, company seats and API keys. The bills come from different places, and none of them says who spent the money or on which feature.

“Who on the team spends the most on AI?”

Spend per person, with the tools each one uses and how that usage is paid for.

“What did this feature cost in AI?”

Every usage entry is attached to the issue the agent was working on, so each feature, bug or task carries its own AI cost.

“Which tool or model is driving the bill?”

Group the same spend by tool, model, provider or cost type, and open any row to see the entries behind it.

From company to session

One total. Every step behind it.

Open a project, then a person, then a feature, and finally the sessions behind it: date, tool, model, tokens and spend. Switch the grouping at any level to see the same money by tool or model.

Specivo · Admin · AI spend Demo data: ACME Waste & Recycling (illustrative)

Company AI spend

Group by

The report in Specivo also groups by provider, cost type and git branch, adds a second “then by” grouping, and filters by person, tool, model, cost type and dates. Each project has its own spend report and AI usage tab, and each issue shows its own AI usage.

Cost per feature

Agent work finds its own issue.

Every usage entry is attached to the issue (feature, bug or task) the agent was working on. Nobody tags anything by hand. Specivo works it out from:

  1. 1The agent’s own Specivo tool calls, on the issue it is working on
  2. 2An environment variable, SPECIVO_ISSUE
  3. 3Commits
  4. 4Branch names
  5. 5The repository
  • Long sessions that switch between tasks are split per issue
  • Research done before the issue existed still counts towards that issue
  • Entries someone has moved by hand are never overwritten
  • Usage with no issue lands on the project, or as unassigned, where people can move it

One session, three issues

Illustrative

Kenji Sato · Claude Code · one long morning session

  • Research before ACMEIT-16 was filed: counted towards ACMEIT-16
  • ACMEIT-16 Migrate shared file repository to NAS, found by branch name
  • ACMEIT-6 Call viewer database locks under load, found by the agent’s Specivo tool call
  • ACMEIT-11 Investigate unified login for all company systems, found by SPECIVO_ISSUE
Moved an entry to another issue by hand? Specivo keeps that choice the next time it attributes usage.
On every issue

What did this feature cost? Open the issue.

Every issue has an AI usage tab. It adds up the agent sessions behind the feature, its sub-issues included: spend, API-equivalent value, tokens, the people who worked on it, the tools and models they used, and how the cost built up week by week.

  • Spend by person, by tool and by model
  • Sub-issues counted, or this issue only
  • Every session behind it, page by page
  • Only what your project role lets you see
The AI usage tab of the Customer portal 1.0 issue: spend, API-equivalent value, tokens and people, spend by person, by tool and by model, and value and spend per week
Demo data: ACME Waste & Recycling (illustrative). The “Customer portal 1.0” feature with its sub-issues.
Every person, every tool

A budget view for the whole team.

See what each person spends on AI, which tools they use, and how that usage is paid for: a personal plan, a company seat or pay-per-use API keys.

  • Everyone sees their own usage on “My AI usage”
  • People see the AI usage their project role allows
  • Individual subscription costs are shown only to those allowed to see them

Spend by person

Demo data: ACME Waste & Recycling (illustrative)
Last 3 months
How it works

Three steps. Nothing to fill in.

01

Install the reporter

ai-usage is an open-source command-line tool (MIT, Python) for Windows, macOS and Linux. It reads the local logs of Claude Code, Codex, Gemini CLI and OpenCode.

02

Work as usual

Its hooks report usage after each agent turn. People and agents work on issues as they always do, and nobody has to log anything.

03

Open the reports

Company, project, issue and personal views, down to every entry. Export to CSV, or read the same data through the REST API and MCP.

terminal
$ ai-usage init
$ ai-usage hooks install --tool all
Reports

Trends you can plan with.

Spend over time, stacked by person, tool, model or project, grouped by day, week or month automatically. Tokens over time, and breakdowns by person, project, model and tool.

Spend over time

Demo data: ACME Waste & Recycling (illustrative) · per week

    Tokens over time

    Demo data: ACME Waste & Recycling (illustrative)

    Input and output

    • Input
    • Output

    Cache

    • Cache read
    • Cache write

    Spend by model

    Demo data: ACME Waste & Recycling (illustrative)

    All kinds of cost

    Plans, seats and API keys. One report.

    Spend is pay-per-use API charges plus each subscription’s fee, spread over the usage that subscription covered. A flat-rate plan and an API key end up in the same report.

    Personal plans

    Presets for the plans people already pay for.

    Claude ProClaude MaxChatGPT PlusChatGPT ProCursor ProGitHub CopilotGoogle AI ProGoogle AI UltraSuperGrok

    Team seats and custom plans

    Business plans such as team seats, and any custom plan. Any currency, with exchange rates.

    SubscriptionAPI (pay-per-use)LocalFree

    Model prices

    A bundled list of prices for more than 300 models, with history. Admins can override any price. Usage without a price is flagged, never silently counted as zero.

    Two extra figures sit next to spend: API-equivalent value, the same tokens at pay-per-use prices (an estimate, not a bill), and leverage, that value divided by the plan fee.

    Privacy & self-hosting

    Numbers leave the laptop. Your work doesn’t.

    The reporter sends usage numbers to your own Specivo, and Specivo runs on your server, air-gapped installs included. The data stays with you.

    What the reporter sends

    • Tokens per model
    • Tokens per session and per day
    • Only to your Specivo server

    The reporter’s API key is stored in the OS keychain.

    What it never sends

    • Prompts
    • Code
    • File contents

    Every piece is open source. Read it yourself.

    The server, the reporter on each laptop and the agent skill are all public code. Anyone can read what is collected, what is sent and where it goes.

    Agent-native

    Agents can report themselves.

    Specivo’s MCP server and an agent skill for Claude Code, Codex and Cursor let agents report their own usage. One line in CLAUDE.md or AGENTS.md turns it on.

    • Ask your agent to register your subscriptions from your invoices
    • Read the usage data through MCP, the REST API or a CSV export

    MCP

    Agents report usage and read it back through Specivo’s MCP server.

    REST API

    The same data for your own scripts and dashboards.

    CSV

    Download any report for a spreadsheet.

    Know where the AI budget goes. On your own server.

    Open source (AGPL). No per-seat pricing. Works with whatever AI your team brings.

    See the live demo Self-host in 15 minutes