Jargon walls
AI vocabulary can make simple ideas sound complicated.
Glossary
The words that matter, in plain English
Tokens, capabilities, routing, fail-over, cache hits, idempotency keys and micros. Everything you need to read a response, an invoice or our docs with confidence.
Plain English
no jargon left unexplained
Tied to the API
terms match the field names
Quick reference
for builders and buyers
The problem
What is slowing you down today
Inconsistent terms
The same idea goes by five names across five tools.
Invoices you cannot read
Billing terms should be as clear as the bill.
Capabilities
Everything you need, nothing you don't
Token
A small chunk of text that models read and write. You are billed for input tokens (your prompt) and output tokens (the response).
Capability
What a model can do, such as chat, code, vision or reasoning. Send capability in the request to let the router choose a model.
Routing and fail-over
Routing picks the lowest-priced enabled model with the capability you asked for. Fail-over retries the request on another eligible model if one is busy or errors.
Cache hit
A response served from cache because the same prompt, model and temperature were seen inside the cache window. cache_hit is true and the price is zero.
Idempotency key
A unique string you send in the Idempotency-Key header so that retries of the same operation are metered once.
Micros
Millionths of a US dollar. price_micros on each response is the exact charge for that call.
Use cases
What teams run on it
01
Reading a response
Know what every field in result means.
02
Explaining a bill
Turn tokens and micros into plain cost.
03
Onboarding teammates
Share one page instead of a long thread.
04
Evaluating vendors
Ask sharper questions about pricing and reliability.
How it works
From sign-up to production
- 1
Skim the terms
Most take one sentence to understand.
- 2
See them in action
Run a prompt in the playground and read the fields.
- 3
Keep it handy
Bookmark this page for your team.
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 free allowance?
A number of tokens each account can use free every month before metered pricing applies.
What is a model_id?
The identifier of a model in our catalog. Send it to use exactly that model.
What is a rate limit?
The maximum requests per minute your account can send. Over it, you get a 429.
What is temperature?
A setting from 0 to 2 that controls how varied the output is. Lower is more focused.
Now put the words to work
Run your first prompt and read every field.