Connect providers
Paste OpenAI, Anthropic, or OpenRouter admin keys. Encrypted in Vault — no SDK, no proxy.
FLUXTAB
One live ledger for OpenAI, Anthropic, and OpenRouter — with staged alerts in chat and an AI spend budget gate (a budget trajectory check, not PR cost prediction).
FLUXTAB / SPEND LEDGER
MONTH TO DATE
$247.86FORECAST
$291.00under capOPENAI 2.4M TOKENS
$142.30ANTHROPIC 1.1M TOKENS
$89.12OPENROUTER 412K TOKENS
$16.44Security / architecture
Fluxtab reads billing APIs. Production traffic still goes straight to OpenAI, Anthropic, and OpenRouter.
Read-only keys · Vault-encrypted · Separate CI token
Usage and billing scope only. Fluxtab never sits in your request path, never sees prompts, and cannot spend on your providers.
Keys are encrypted in Supabase Vault the moment they arrive, decrypted only in memory during a sync, never sent to the browser.
Pipeline tokens can read spend numbers. They cannot read provider keys. Revoke either independently from Settings.
What you get
Four moves, one story — from the ledger to the workflow, not a flat feature dump.
01See
OpenAI, Anthropic, and OpenRouter on one timeline — with a month-end pace so you know where the cycle is heading.
02Explain
Project splits, a spend heatmap, and same-day spike detection — so a jump is a cause, not a mystery.
03Control
One monthly threshold. Staged alerts at 50 · 80 · 100 · 120%. The AI spend budget gate fails the deploy when the account is already over.
04Integrate
Read-only API, webhooks, Slack / Discord / Teams, and shared team seats — no SDK in your request path.
READY?
Connect a provider in under two minutes. Free to start — upgrade when you need the ladder, the gate, or the team seats.
START TRACKING FREESETUP
Three moves from scattered billing portals to a live cost signal your team — and your pipeline — can act on.
Paste OpenAI, Anthropic, or OpenRouter admin keys. Encrypted in Vault — no SDK, no proxy.
Daily sync into one timeline, with forecast, heatmap, and project splits as data lands.
Staged alerts in chat. CI runs a budget trajectory check before deploy.
Fluxtab
Usage receipt
2026-08-08UTC
*** thank you for building ***
03 / Pricing
Four tiers from the same catalog that powers billing. Start free, then unlock providers, history, alerts, the AI spend budget gate, and seats as you grow.
2 months freeon yearly billing
For solo builders getting their first read on API spend.
See your AI spend.
Unlimited providers plus the full early-warning stack.
Know when it's drifting.
AI spend budget gate, public API, and email alert channels.
Control AI spend across your team.
Everything in Business with seats for your org.
Put AI spend into the engineering workflow.
Prices in INR via Razorpay. Yearly is billed once and includes 2 months free. Open an account ↗
04 / FIELD NOTES
The useful answers, without a sales call or a forty-page security deck — including what Fluxtab can and can't stop.
No. Fluxtab reads billing and usage data from provider APIs. Your production requests keep going directly to OpenAI, Anthropic, OpenRouter, or whichever provider you use.
You can connect one provider, view 7 days of history, and get a Slack alert when you cross your monthly threshold. It is enough to see whether the workflow fits before upgrading.
You set one monthly threshold. On Free you get a single alert at 100%. On Pro and above, Fluxtab warns you in stages — 50%, 80%, 100%, and 120% — and separately flags any day where spend spikes well above your recent average. Alerts deliver to Slack, Discord, Microsoft Teams, email, or a raw webhook, depending on plan.
Type /fluxtab spend in Slack or Discord and the bot replies with month-to-date, forecast, and per-provider spend — private to you, no dashboard tab needed. Available on Pro and above.
Your pipeline calls a single endpoint with a read-only bearer token. This checks your account's budget trajectory as of this PR — it does not yet calculate the specific cost impact of the code change itself. If you're already over budget it returns a failure your CI treats as a failed step. Available on Business and Team plans, along with the public summary API.
On the Team plan you can invite up to 5 teammates to the same cost ledger with their own logins. Everyone sees the same numbers; nobody shares provider keys over chat.
OpenAI and Anthropic history follows your plan's range (7/30/90/full days). OpenRouter is capped at the last 30 days regardless of plan, due to a limitation in OpenRouter's own usage API — we'll lift this if OpenRouter's longer-range analytics endpoint becomes stable.
Keys are handled server-side and stored through the app's Supabase-backed secret workflow. They are not exposed to the browser, and public API tokens can never read them.
Not at the provider level. Connected keys are read-only by design — that's what makes them safe to hand over — so Fluxtab alerts you rather than blocking a request. The one place it can hard-stop something is your own pipeline: the AI spend budget gate fails the build when the account is already over budget.
Not reliably. As of 2026, OpenAI's account budget setting notifies but doesn't hard-stop requests, and Anthropic has no equivalent cap for standard API keys. Treat Fluxtab and your provider's own settings as early warning, not a guarantee.
Your last 7 days of usage, projected to the end of the billing cycle. It's an estimate, not a guarantee — shown only once you have at least 3 days of data.
On Free, provider usage syncs about every 6 hours. On paid plans, sync runs every hour. This isn't live monitoring — a spike shows up same-day, not same-minute. You can always tap Sync now on the dashboard.