Usage metering platform
One eye for every API.
Everything you build on now runs on a meter — tokens, credits, calls, rate limits. You shouldn't need ten browser tabs to know where they all stand. Argus reads every meter onto one page.
Early access. No spam — just the launch.
One place, instead of a dozen dashboards.
Consumption pricing put a meter on everything you use — OpenRouter credits, OpenAI tokens, Firecrawl calls, per-minute rate limits. Each sits behind its own login, so you check them one at a time, hoping nothing ran dry overnight. Argus reads every meter into a single view: where you stand, how fast it's moving, and when a vendor is about to cut you off.
Three ways in
Use one or all three. Poll for authoritative totals, push for real-time detail, MCP for agents.
Poll with an API key
Add a provider key and Argus pulls balances and usage on a schedule — authoritative totals and remaining credits, zero code.
Push with the SDK
Drop @argus/sdk into your app's egress. It captures per-call spend and the rate-limit headers no billing endpoint exposes — for any provider.
Query over MCP
Point your agents at the Argus MCP server so they keep tabs on spend and rate-limit headroom, and route around trouble.
Tracks the whole stack
…and any other API you call.
Burn rate & days-to-zero
Know exactly when each credit balance runs dry — before it stops your app mid-run.
Rate-limit headroom
Track RPM/TPM ceilings and 429 rates from your own traffic. See the wall before you hit it.
Alerts that matter
Low balance, days-to-zero, spend spikes, projected overage, error-rate — by email.
The long tail
Not just LLMs. Data, enrichment, and scraping APIs too — the vendors nobody else tracks.
Agent-ready
A read-only MCP server so your agents watch usage and optimize spend automatically.
Yours, encrypted
Multi-tenant from day one. Provider credentials are AES-256 encrypted at rest — never logged.
Stop guessing what you're spending.
Join the waitlist and be first in when Argus opens up.