One endpoint for every model.
An OpenAI-compatible API for open and frontier models. Point any SDK at
/v1, use a single key, and pay per token at transparent rates —
no seats, no surcharges, no markup. Swap models in one line.
curl https://api.opexai.com/v1/chat \ -H "Authorization: Bearer opex_…" \ -d '{ "model": "llama-3.3-70b", "messages": [...] }'
Drop-in in three minutes.
If your client speaks OpenAI, it already works. Change the base URL, add a key, pick a model — that's the whole integration.
Point the SDK
Any OpenAI client works unchanged. Swap the base URL and drop in an opex key.
base_url = "https://api.opexai.com/v1" api_key = "opex_…"
Pick any model
One key unlocks the whole catalog — chat, reasoning, vision, embeddings. Switch by changing one string.
POST /v1/chat/completions POST /v1/embeddings
Pay per token
Billed transparently per million tokens. Prepaid credits or monthly invoice — your call.
cost = tokens × rate / 1M
A router, not a lock-in.
Standards-first by design. Bring any OpenAI SDK, leave any time. The value is the routing, caching, and metering underneath.
OpenAI-compatible
Same shapes, same paths, same streaming. The Python and JS SDKs work with zero changes — just a new base URL.
Runs at the edge
Requests are served close to your users with low latency, automatic caching, and built-in rate limiting.
Transparent pricing
Every model lists its per-million-token rate up front. No seats, no tiers, no surcharge on top.
Per-key metering
Issue as many keys as you like. Usage is tracked per key, with live balances and exports for invoicing.
Pricing, plain.
Rates per million tokens. The same number you see here is the number you pay — input and output priced separately.
| Model | Output / 1M |
|---|
Ship in an afternoon.
Not a quarter.
Generate a key, change your base URL, and you're routing to every model on the catalog. Free daily allowance to get started — no card required.
Open the console →