OpenAI-Compatible Endpoint

EpsimoAI provides an OpenAI-compatible API at /public/openai/v1/chat/completions. This allows you to use any tool or SDK that supports the OpenAI API format — including the official OpenAI Python SDK, Typebot, Zapier, and other no-code platforms — to interact with your EpsimoAI assistants.

How It Works

The endpoint mimics OpenAI's /v1/chat/completions interface. The per-assistant public token acts as the API key and identifies which assistant handles the request. No separate assistant_id or model selection is needed — the backend resolves the assistant from the token.

Step 1: Configure Your Assistant for Public Access

Before using the OpenAI-compatible endpoint, configure your assistant with a public token.

Via the Frontend

  1. Go to https://chat.epsimoagents.com
  2. Navigate to your assistant's settings
  3. Set Access Mode to public_token
  4. Set Public Token to a secret string (e.g., pubkey_mybot_abc123)
  5. Save

Via the API

curl -X PATCH https://backend.epsimoagents.com/assistants/<assistant-id> \
  -H "Authorization: Bearer $JWT" \
  -H "Content-Type: application/json" \
  -d '{
    "config": {
      "configurable": {
        "type==agent/access_mode": "public_token",
        "type==agent/public_token": "pubkey_mybot_abc123"
      }
    }
  }'

See Public Assistants for full details on access modes.

Step 2: Call the Endpoint

Endpoint

POST https://backend.epsimoagents.com/public/openai/v1/chat/completions

Headers

Header Value
Authorization Bearer <your-public-token>
Content-Type application/json

Request Body

Standard OpenAI chat completions format:

{
  "model": "gpt-4o",
  "messages": [
    {"role": "user", "content": "Hello, what can you help me with?"}
  ],
  "stream": false
}

Note: The model field is accepted for compatibility but ignored — the assistant uses whatever model is configured in its settings.

Calling with curl

Non-streaming

curl https://backend.epsimoagents.com/public/openai/v1/chat/completions \
  -H "Authorization: Bearer pubkey_mybot_abc123" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gpt-4o",
    "messages": [{"role": "user", "content": "What is EpsimoAI?"}],
    "stream": false
  }'

Response:

{
  "id": "chatcmpl-...",
  "object": "chat.completion",
  "choices": [
    {
      "index": 0,
      "message": {
        "role": "assistant",
        "content": "EpsimoAI is..."
      },
      "finish_reason": "stop"
    }
  ]
}

Streaming

curl -N https://backend.epsimoagents.com/public/openai/v1/chat/completions \
  -H "Authorization: Bearer pubkey_mybot_abc123" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gpt-4o",
    "messages": [{"role": "user", "content": "Tell me a story"}],
    "stream": true
  }'

Streaming response (SSE format):

data: {"id":"chatcmpl-...","object":"chat.completion.chunk","choices":[{"index":0,"delta":{"role":"assistant","content":"Once"},"finish_reason":null}]}

data: {"id":"chatcmpl-...","object":"chat.completion.chunk","choices":[{"index":0,"delta":{"content":" upon"},"finish_reason":null}]}

data: [DONE]

Configuring in Typebot

Typebot natively supports OpenAI-compatible APIs. To connect your EpsimoAI assistant:

  1. In your Typebot flow, add an OpenAI block (or "AI" block)
  2. Configure the connection:
  3. Base URL: https://backend.epsimoagents.com/public/openai/v1
  4. API Key: pubkey_mybot_abc123 (your assistant's public token)
  5. Set the model to any value (e.g., gpt-4o) — it will be ignored by EpsimoAI
  6. Configure your messages as usual in Typebot

Typebot will automatically append /chat/completions to the Base URL and send requests in the standard OpenAI format.

Using the OpenAI Python SDK

Point the official OpenAI Python SDK at EpsimoAI by overriding the base_url:

from openai import OpenAI

client = OpenAI(
    api_key="pubkey_mybot_abc123",
    base_url="https://backend.epsimoagents.com/public/openai/v1",
)

# Non-streaming
response = client.chat.completions.create(
    model="gpt-4o",  # ignored by EpsimoAI, but required by the SDK
    messages=[
        {"role": "user", "content": "Hello! What can you do?"}
    ],
)
print(response.choices[0].message.content)

Streaming with the Python SDK

from openai import OpenAI

client = OpenAI(
    api_key="pubkey_mybot_abc123",
    base_url="https://backend.epsimoagents.com/public/openai/v1",
)

stream = client.chat.completions.create(
    model="gpt-4o",
    messages=[{"role": "user", "content": "Explain quantum computing simply"}],
    stream=True,
)

for chunk in stream:
    if chunk.choices[0].delta.content:
        print(chunk.choices[0].delta.content, end="", flush=True)
print()

Multi-turn Conversations

Pass the full conversation history each time:

messages = [{"role": "user", "content": "My name is Alice"}]

response = client.chat.completions.create(model="gpt-4o", messages=messages)
assistant_reply = response.choices[0].message.content
messages.append({"role": "assistant", "content": assistant_reply})

messages.append({"role": "user", "content": "What's my name?"})
response = client.chat.completions.create(model="gpt-4o", messages=messages)
print(response.choices[0].message.content)  # "Your name is Alice"

Using the OpenAI Node.js SDK

import OpenAI from "openai";

const client = new OpenAI({
  apiKey: "pubkey_mybot_abc123",
  baseURL: "https://backend.epsimoagents.com/public/openai/v1",
});

const response = await client.chat.completions.create({
  model: "gpt-4o",
  messages: [{ role: "user", content: "Hello from Node.js!" }],
});

console.log(response.choices[0].message.content);

Differences from the Real OpenAI API

Feature OpenAI EpsimoAI
model parameter Selects the model Ignored (assistant config determines model)
temperature, top_p Supported Passed through if underlying model supports them
functions / tools Supported Not supported via this endpoint
n (multiple completions) Supported Not supported
Token usage in response Always present May not be present
Rate limits Per-API-key No built-in rate limiting

Troubleshooting

Error Cause Fix
401 Missing authentication token No Authorization header Add Authorization: Bearer <token>
401 Invalid token Token doesn't match any assistant Check the token value matches config
403 Not configured for public access access_mode is authenticated Set access_mode to public_token
404 Assistant not found No assistant has this token Verify token is set in assistant config