Skip to content
Log in

Developers

Build with Atom.

The Incbury API serves Atom through an OpenAI-compatible chat completions endpoint. Bring an existing SDK, change the base URL and pay per token from prepaid credits.

Get an API keyAPI pricing

Quickstart

Your first request.

  1. 1

    Create a key

    Sign in to the developer platform and create an API key. It is shown once, so store it safely.

  2. 2

    Add credits

    Requests are paid from prepaid credits. You can see your balance and usage on the platform.

  3. 3

    Call the model

    Send a chat completions request to the model atom with your key as a bearer token.

curl https://api.incbury.co.uk/v1/chat/completions \
  -H "Authorization: Bearer $INCBURY_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "atom",
    "messages": [{"role": "user", "content": "Summarise the water cycle in two sentences."}],
    "reasoning_effort": "none"
  }'

Works with the official OpenAI SDKs by setting the base URL. Examples use the environment variable INCBURY_API_KEY.

SDKs

Use the SDK you already know.

The incbury packages are thin wrappers around the official OpenAI SDKs with the base URL and key already set. Every OpenAI-compatible tool works too.

Python

pip install incbury

from incbury import Incbury

client = Incbury()  # reads INCBURY_API_KEY
reply = client.chat.completions.create(
    model="atom",
    messages=[{"role": "user", "content": "Hello, Atom!"}],
)
print(reply.choices[0].message.content)

TypeScript and JavaScript

npm install incbury openai

import Incbury from "incbury";

const client = new Incbury(); // reads INCBURY_API_KEY
const reply = await client.chat.completions.create({
  model: "atom",
  messages: [{ role: "user", content: "Hello, Atom!" }],
});
console.log(reply.choices[0].message.content);

Or keep the OpenAI SDK and change the base URL

from openai import OpenAI

client = OpenAI(base_url="https://api.incbury.co.uk/v1", api_key=os.environ["INCBURY_API_KEY"])

The model

What you are calling.

Model ID
atom
Base URL
https://api.incbury.co.uk/v1
Authentication
Authorization: Bearer atom_…
Context window
16,384 tokens
Maximum output
4,096 tokens
Supports
Chat, reasoning effort, streaming, function calling, structured outputs, image input, logprobs, embeddings, fine-tuning, batches and evals
Not supported
Realtime voice sessions (coming soon)

The API serves the current Atom release. See its evaluations and limitations.

Reference

Endpoints and parameters.

Endpoints

  • POST /v1/chat/completionsCreate a chat completion, streamed or not
  • POST /v1/responsesStateful responses with built-in web search, file search, code interpreter and image tools
  • GET /v1/modelsList available models
  • GET /v1/models/{id}Get one model's details and limits
  • POST /v1/embeddingsTurn text into vectors
  • POST /v1/moderationsClassify text for harmful content
  • POST /v1/images/generationsGenerate and edit images
  • POST /v1/audio/transcriptionsTranscribe audio to text
  • POST /v1/audio/speechTurn text into natural speech
  • POST /v1/filesUpload files for batches, fine-tuning, evals and search
  • POST /v1/vector_storesIndex files for retrieval and file search
  • POST /v1/batchesRun large jobs asynchronously
  • POST /v1/fine_tuning/jobsTrain a LoRA adapter on your own examples
  • POST /v1/evalsGrade models and fine-tunes against your dataset
  • POST /v1/tokenizeCount tokens for a piece of text
  • GET /v1/tokenizer/{model}Download the tokenizer

Chat completion parameters

  • messages
  • stream
  • stream_options.include_usage
  • max_tokens
  • max_completion_tokens
  • temperature
  • top_p
  • stop
  • n
  • logprobs
  • top_logprobs
  • reasoning_effort
  • response_format
  • tools
  • tool_choice
  • user

Messages can use the system, developer, user, assistant and tool roles. Up to 64 function tools per request. Send an Idempotency-Key header to retry safely. Requests and responses pass through moderation.

Pricing

Pay per token.

Prices are per million tokens, in pounds sterling, paid from prepaid credits.

Incbury API prices per million tokens
TokensPrice per million
Input£0.20
Cached input£0.05
Output£0.80

Images, audio, video, tools and training

Images, audio, video, tools and training
ProductPrice per million
Embeddings£0.02 per million tokens
Image generation£0.03 per image
Image editing£0.04 per image
Video generation£0.06 per second
Speech to text£0.004 per minute
Text to speech£0.01 per 1,000 characters
Web search tool£0.005 per call
File search tool£0.002 per call
Code interpreter tool£0.003 per call
Vector storage£0.08 per GB per day
Fine-tuning£0.50 per million trained tokens
  • Fine-tuning jobs have a minimum charge of £0.50. Fine-tuned models are served at their base model's token prices.
  • When your credits run out, requests return HTTP 402 insufficient_credits. Each response reports your remaining balance in the x-credits-remaining header.
  • Each key allows 60 requests a minute by default, and you can set a different limit per key.
  • Incbury Code is paid from the same credits.
  • Incbury Chat has its own plans. Compare chat plans

Resources

Everything else you need.

Developer platform

Keys, usage, credits and request logs in one place.

API keys

Create, limit and revoke keys.

Playground

Try prompts and settings before you write code.

Tokenizer

See how Atom splits your text into tokens.

Credits and billing

Top up credits and download invoices.

Changelog

Product and API changes as they ship.

Start building with Atom.

Create a key, add credits and send your first request in a few minutes.

Try Incbury CodeGet an API key