Every AI interaction consumes tokens. Every token has a story.
One OpenAI-compatible API for 106 frontier models, routed intelligently across 25 providers. Watch every token like an asset — because it is one.
AIT / USD
—
- API-100 index
- —
- —
- Frontier models
- —
- —
- Tokens routed
- —
- …and counting
- Frontier models
- —
- One OpenAI-compatible API
- Providers
- —
- Routed per request
- Tokens processed
- —
- …and counting
- Free models
- —
- :free variants, $0 forever
- API-100 index
- —
- —
Model marketplace
Find your model
Live pricing and usage for every model on the gateway — sorted the way you actually shop.
Developer quickstart
Ship your first request in 60 seconds
You don't need a migration plan — if it speaks OpenAI, it already speaks aiToken.
- 1
Create a key
One click in the console gives you a sk-a1tk- key. No sales call, no card.
- 2
Pick a model
Any slug from the marketplace — GPT-5, Claude, Gemini, Llama — same schema.
- 3
Make one call
Point your SDK at our base URL. Your first tokens stream back in seconds.
curl https://aitoken.com/api/v1/chat/completions \
-H "Content-Type: application/json" \
-H "Authorization: Bearer sk-a1tk-xxxxxxxxxxxxxxxx" \
-d '{
"model": "openai/gpt-4o",
"messages": [
{ "role": "user", "content": "Introduce yourself in one sentence" }
]
}'Your first 100k tokens are on us — no credit card.
Why teams switch
A gateway that behaves like a product, not a proxy
Everything below ships today — no beta flags, no enterprise-only fine print.
Smart Routing
Price-first, performance-first, and failover-first strategies pick the best upstream per request, in real time.
Token Cost Simulator
Tune the input/output token mix with sliders and instantly rank every model by real-world cost.
Usage Analytics Dashboard
Break down spend, tokens, requests, latency, and cache hit rate by model and API key.
Multi-level Failover
Unhealthy upstreams automatically fall back to backup providers — high success rates out of the box.
Multi-Model Comparison
Compare up to 4 models side by side — specs, cost curves, and capability radar. No more guesswork.
Free & Open-source Filters
One click to surface :free variants and open-weight models — great for tinkerers and self-hosting.
Live Platform Activity
LIVEAI asset markets
One terminal, every AI asset
Spot, model tokens, stocks, derivatives, hardware and compute — all quoted in the same terminal.
25 Global Providers · One Gateway
Top 12 by market share · click a card for provider details and all its models
aiToken Price Index
30-day trajectory of the network-wide weighted output price · first day = 100 (API-100 baseline)
The aiToken Economy
Beyond the Gateway — One Economy, Four Pillars
From trading and capital to compute and research — every pillar of the AI economy lives under one roof.
Newsroom
Latest from the AI Newsroom
Headlines, funding and research — curated around the clock.
Primary Market
Capital Radar
Where private capital is flowing across AI unicorns — latest disclosed rounds, live from the unicorn board.
Built for builders
Teams don't switch for the buzzwords
They switch for the invoice. A few words from people who did.
“We cut our LLM spend 38% in the first month — the cost simulator found two models doing the same job at a third of the price, and price-first routing keeps it that way.”
“One base-URL swap turned 25 provider accounts into a single line item. Failover re-routed mid-launch-week outage before our on-call even opened the dashboard.”
“I prototype on :free models all day, then flip to GPT-5 for production without touching a line of code. My AI bill is still under $200 a month.”
Frequently Asked Questions
Quick answers on billing, integration, and data privacy
Start Building Today
Free credits on sign-up, your first API key in 30 seconds, and unlimited access to the Chat Playground.