Days of setup
Most platforms want projects, regions and deployments before you see a single token.
Developers
Your first response in minutes
Sign up, create a key, send one POST request. That is the whole integration. Everything else, routing, fail-over, caching and metering, happens on our side.
One request
is the whole integration
Plain HTTP
works from any language
Free to start
monthly token allowance included
The problem
What is slowing you down today
Client lock-in
You should not need a special package to send a prompt and read a response.
Unknown costs while you build
Experiments should tell you what they cost as you run them.
Capabilities
Everything you need, nothing you don't
Create a key
In the console, open Settings and create an API key. Copy it immediately: it is shown once and stored only as a hash.
Send a prompt
POST https://<your-host>/api/ainf/inference with the header Authorization: Bearer <your key>, Content-Type: application/json and a body of {"prompt": "Write a haiku about latency"}.
Read the result
The response is {"ok": true, "result": {...}} with response, model_id, model_name, input_tokens, output_tokens, total_tokens, latency_ms, cache_hit, price_micros and status.
Pick a capability
Add "capability": "code" (or vision, reasoning and others in the catalog). Without it, chat is assumed and the lowest-priced enabled chat model answers.
Or name a model
Add "model_id" with an id from the model catalog to use exactly that model.
Tune the output
Optional temperature (0 to 2, default 0.7) and max_tokens (1 to 8192, default 1024) control the generation.
Use cases
What teams run on it
01
curl from a terminal
The fastest way to see a real response and its price.
02
A backend service
Any HTTP client in any language works. Keep the key server-side.
03
An automation tool
Configure a generic HTTP step with the URL, header and JSON body.
04
The playground
No code at all: run prompts from the console with your signed-in session.
How it works
From sign-up to production
- 1
Sign up
Create an account. Your free monthly token allowance starts immediately.
- 2
Create a key
Settings, API keys, create. Store it in your secret manager.
- 3
POST /api/ainf/inference
Send {"prompt": "..."} with your bearer key and read result.response.
Control
Built so you stay in charge
Keys you control
API keys are shown once at creation and stored only as a hash. Revoke any key instantly from Settings.
Spend you can see
Every request carries its own receipt: tokens in, tokens out, latency and price. Nothing is billed that you cannot trace.
Limits you set
Per-account rate limits and a free monthly allowance keep experiments safe before they become production.
Data you decide on
You choose exactly what you send. Repeat prompts can be served from your cache, and cache hits are never billed.
For developers
One endpoint. Every model.
Send a prompt and a capability, or name a model. One bearer key, one small JSON body, and a full receipt on every response.
curl https://YOUR-DOMAIN/api/ainf/inference \
-H "Authorization: Bearer $AI_INFERENCE_KEY" \
-H "Content-Type: application/json" \
-H "Idempotency-Key: order-1042" \
-d '{"prompt":"Summarise this ticket in one line","capability":"chat"}'Questions, answered
What is the minimum request body?
{"prompt": "your text"}. The prompt must be a non-empty string. Every other field is optional.
How do I authenticate?
Send Authorization: Bearer <your API key>. Requests from the signed-in console use your session instead.
Do you publish client packages?
No package is needed. The API is plain HTTPS and JSON, so any HTTP client works.
What does price_micros mean?
The exact price of the call in millionths of a US dollar. Cached responses and calls inside your free allowance report zero.
Make your first call now
Create a key and send one request. The free allowance covers it.