Developer API

BotHub exposes an OpenAI-compatible HTTP API. The balance in your BotHub account can be spent directly from your own programs and scripts — or by an AI assistant you hand the credentials to.

Because the API is OpenAI-compatible, anything that accepts a custom base URL and API key works out of the box: the official OpenAI SDKs, LangChain, most desktop clients, and coding assistants like Claude or Codex.

Base URL and authentication

https://bothub-api.bookab.info/v1

Send your API key with every request:

Authorization: Bearer bhk_live_xxxxxxxx

With an OpenAI SDK, only two parameters change:

from openai import OpenAI

client = OpenAI(
    base_url="https://bothub-api.bookab.info/v1",
    api_key="bhk_live_xxxxxxxx",
)

resp = client.chat.completions.create(
    model="gpt-5.4",
    messages=[{"role": "user", "content": "Hello"}],
)
print(resp.choices[0].message.content)

import OpenAI from 'openai'

const client = new OpenAI({
  baseURL: 'https://bothub-api.bookab.info/v1',
  apiKey: 'bhk_live_xxxxxxxx',
})

const resp = await client.chat.completions.create({
  model: 'gpt-5.4',
  messages: [{ role: 'user', content: 'Hello' }],
})
console.log(resp.choices[0].message.content)

curl https://bothub-api.bookab.info/v1/chat/completions \
  -H "Authorization: Bearer bhk_live_xxxxxxxx" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gpt-5.4",
    "messages": [{"role": "user", "content": "Hello"}]
  }'

Start by listing the models

Don't guess model names. One call tells you which models your key can use and what each one is:

curl https://bothub-api.bookab.info/v1/models \
  -H "Authorization: Bearer bhk_live_xxxxxxxx"

{
  "object": "list",
  "data": [
    {
      "id": "gpt-5.4",
      "object": "model",
      "owned_by": "bothub",
      "kind": "chat",
      "endpoints": ["/v1/chat/completions", "/v1/responses"]
    },
    {
      "id": "gpt-image-2.5",
      "object": "model",
      "owned_by": "bothub",
      "kind": "image_generation",
      "endpoints": ["/v1/images/generations"]
    }
  ]
}

kind and endpoints are BotHub extensions (OpenAI SDKs ignore unknown fields): kind tells you whether it is a chat, image, video or embedding model, and endpoints tells you exactly where to send the request.

What you can do

Not available yet: speech synthesis (/v1/audio/speech returns not_implemented); reranking has no public endpoint.

Handing the base URL and key to an AI

If you want an AI assistant (Claude, Codex, some agent client) to call the API for you, this is all it needs:

The BotHub API is OpenAI-compatible.
Base URL: https://bothub-api.bookab.info/v1
API Key: bhk_live_xxxxxxxx

Start with GET /v1/models to see what is available — each entry's `kind` and
`endpoints` tell you what it is and where to send requests. For images use
POST /v1/images/generations; it returns data[0].b64_json synchronously.
Full docs: https://bothub.bookab.info/en/api/overview

Getting an API key

Create one yourself inside the BotHub app — all three platforms have it:

That screen copies the base URL, creates and revokes keys, and shows each key's request count and spend over the last 7 days.

New keys carry every scope by default (chat, images, video, embeddings), so they work right away.

Next

• Chat and embeddings — streaming, tool calls, vision input
• Images and video — synchronous images, async tasks, editing
• Errors, limits and billing — what each error code means and how to debug