AI API (OpenAI-Compatible)

AACFlow exposes an OpenAI-compatible AI API at /api/v1/ai. Any client or SDK that targets the OpenAI API can be pointed at AACFlow by changing the base URL and swapping the API key.

All AI API requests deduct from your AACFlow credit balance. A small deposit is reserved at the start of each request and reconciled against the actual model cost once the response is complete.

Authentication

Pass your AACFlow API key in either the x-api-key header (native) or the Authorization: Bearer <key> header (OpenAI SDK compatible).

Authorization: Bearer aacf_your_api_key_here

Base URL

https://www.aacflow.io/api/v1/ai

For local development:

http://localhost:4000/api/v1/ai

Chat Completions

POST /api/v1/ai/chat/completions

Generate a model response for a conversation. Request and response schemas are compatible with the OpenAI Chat Completions API.

Request body

Field	Type	Required	Description
`model`	`string`	Yes	Model identifier, e.g. `gpt-4o-mini`, `claude-sonnet-4-6`
`messages`	`Message[]`	Yes	Conversation history. At least one message required
`temperature`	`number`	No	Sampling temperature 0–2
`max_tokens`	`integer`	No	Maximum tokens to generate
`stream`	`boolean`	No	Stream response as SSE (default `false`)
`top_p`	`number`	No	Nucleus sampling
`stop`	`string \| string[]`	No	Stop sequences
`tools`	`Tool[]`	No	Function-calling tools
`tool_choice`	`string \| object`	No	`"auto"`, `"none"`, `"required"`, or named function
`response_format`	`object`	No	`\{ "type": "json_object" \}` or JSON schema

Response headers

Header	Description
`X-Request-Id`	Unique request identifier for support
`X-Credits-Used`	Credits deducted for this request (USD, 6 decimal places)
`X-Credits-Remaining`	Credit balance after this request

Example — non-streaming

curl -X POST https://www.aacflow.io/api/v1/ai/chat/completions \
  -H "Authorization: Bearer aacf_your_key" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gpt-4o-mini",
    "messages": [
      { "role": "user", "content": "Summarise the top 3 risks in this contract." }
    ],
    "temperature": 0.2
  }'

from openai import OpenAI

client = OpenAI(
    api_key="aacf_your_key",
    base_url="https://www.aacflow.io/api/v1/ai",
)

response = client.chat.completions.create(
    model="gpt-4o-mini",
    messages=[{"role": "user", "content": "Hello!"}],
)
print(response.choices[0].message.content)

import OpenAI from 'openai'

const client = new OpenAI({
  apiKey: 'aacf_your_key',
  baseURL: 'https://www.aacflow.io/api/v1/ai',
})

const response = await client.chat.completions.create({
  model: 'gpt-4o-mini',
  messages: [{ role: 'user', content: 'Hello!' }],
})
console.log(response.choices[0].message.content)

Example response

{
  "id": "chatcmpl-a1b2c3d4e5f6",
  "object": "chat.completion",
  "created": 1720000000,
  "model": "gpt-4o-mini",
  "choices": [
    {
      "index": 0,
      "message": {
        "role": "assistant",
        "content": "Hello! How can I help you today?"
      },
      "finish_reason": "stop"
    }
  ],
  "usage": {
    "prompt_tokens": 9,
    "completion_tokens": 12,
    "total_tokens": 21
  }
}

Streaming

Set "stream": true to receive a text/event-stream response. Each chunk is a data: {...} SSE line with a partial ChatCompletionChunk object, terminated by data: [DONE].

curl -X POST https://www.aacflow.io/api/v1/ai/chat/completions \
  -H "Authorization: Bearer aacf_your_key" \
  -H "Content-Type: application/json" \
  -d '{"model":"gpt-4o-mini","messages":[{"role":"user","content":"Count to 5"}],"stream":true}'

List Models

GET /api/v1/ai/models

Returns the list of model identifiers available through the AI API.

curl https://www.aacflow.io/api/v1/ai/models \
  -H "Authorization: Bearer aacf_your_key"

Error responses

All errors follow the OpenAI error envelope format:

{
  "error": {
    "message": "Human-readable description",
    "type": "invalid_request_error",
    "code": "model_not_found"
  }
}

HTTP status	`type`	`code`	Cause
401	`invalid_request_error`	`invalid_api_key`	Missing or invalid API key
400	`invalid_request_error`	—	Malformed request body
402	`insufficient_quota`	`insufficient_credits`	Credit balance too low
404	`invalid_request_error`	`model_not_found`	Model not supported
500	`server_error`	—	Internal error

Supported models

AACFlow routes requests to the underlying provider based on the model identifier. Any model listed on the Models page can be used here. The identifier is the same slug used elsewhere in the platform.

Not all models support all features (e.g. tool calling, structured output). Capabilities are documented on the individual model pages.

Using with LangChain / LlamaIndex

Point the OpenAI-compatible integration at the AACFlow base URL:

from langchain_openai import ChatOpenAI

llm = ChatOpenAI(
    model="claude-sonnet-4-6",
    openai_api_key="aacf_your_key",
    openai_api_base="https://www.aacflow.io/api/v1/ai",
)

from llama_index.llms.openai import OpenAI

llm = OpenAI(
    model="gpt-4o-mini",
    api_key="aacf_your_key",
    api_base="https://www.aacflow.io/api/v1/ai",
)

AI API (OpenAI-Compatible)

On this page