What is this costing us?
Live spend by developer, model, repo and tool — including subscription seats, not just API tokens. "The migration cost £1.4k in AI this month" becomes a sentence you can say out loud.
Headroom shows what Claude Code and friends cost per developer, repo and model — and which plan each seat should actually be on. Read-only, out of your request path, running in five minutes.
Request early accessFrom the author of ksred.com — the Claude Code pricing and cost-tracking guides.
Dashboards report. Headroom recommends. It exists for the engineering manager who signs the renewal, not the SRE who tunes the collector.
Live spend by developer, model, repo and tool — including subscription seats, not just API tokens. "The migration cost £1.4k in AI this month" becomes a sentence you can say out loud.
Utilisation per seat, flagged before the renewal date, not after. One ghost seat at $125 a month pays for Headroom several times over.
A per-developer recommendation — API, Pro, Max 5×, Max 20× or a Team seat — built from real volume and limit-hit data. Export it, forward it to finance, done.
Claude Code already emits OpenTelemetry. Headroom is just the endpoint that turns it into answers — no Grafana, no PromQL, no agent to deploy.
Two lines in your team’s shell profile, or one MDM policy.
Token counts, costs, models, session and repo attributes. Never prompts, never code, never in the request path. Nothing changes about how anyone works, and nothing breaks if we go down.
Spend meters per seat, budgets with clip alerts in Slack, and the renewal report — from the first day of data. No dashboard to build.
Headroom watches real usage against every plan’s economics and limits, then tells you — per developer — what each seat should be on. The output is a one-page report your finance team actually reads.
Set a monthly budget per team or repo. Spikes, out-of-hours agent runs and unfamiliar repos raise a clip alert in Slack before the invoice does.
Spend rolls up by repository and project from the telemetry’s own attributes. Cost per initiative, without manual tagging.
Metrics only — no prompts, no code, no keystrokes. Aggregate views by default and no productivity leaderboards. Works-council friendly.
Claude Code first. Codex, Copilot and Gemini CLI importers follow — one neutral view no single vendor will ever give you.
Paste an endpoint, watch the meters move.
The same single Go binary, plus a licence key.
| Edition | Price | Includes |
|---|---|---|
| HostedFree | £0 | One seat, 30-day history |
| HostedTeam | £6 / seat / mo | Minimum 5 seats, unlimited history |
| HostedBusiness | £12 / seat / mo | SSO, API access, priority support |
| Self-hostedTeam | £490 / year | Up to 25 seats, updates included |
| Self-hostedScale | £990 / year | Up to 100 seats |
A premium AI coding seat runs $100–125 a month. Headroom finding one unused seat, or one developer on the wrong plan, covers a year of either edition.
Assumes £100 per developer per month for a premium AI coding seat. Self-hosted works out cheaper above roughly 7 developers.
Early access opens to small teams running Claude Code today. Design partners get the hosted Team plan free while we build the renewal report together.
No newsletter, no drip sequence. One email when your invite is ready.