Skip to Content
Model GuideIntegrationsOpenAI SDK

OpenAI SDK

For server-side applications using the official OpenAI SDK, change the client address and supply a Fenno key. Also select an available model ID: a model from your previous provider is not automatically available through Fenno.

Environment variables

Replace both placeholders. Start with a Chat Completions model for FENNO_MODEL; check Responses support separately before using that example.

export FENNO_API_KEY="<YOUR_FENNO_API_KEY>"
export FENNO_MODEL="<YOUR_MODEL_ID>"

Python

Install the SDK and run the example in the same terminal:

python -m pip install --upgrade openai
example.py
import os
from openai import OpenAI
 
client = OpenAI(
    base_url="https://api.fenno.ai/v1",
    api_key=os.environ["FENNO_API_KEY"],
)
 
response = client.chat.completions.create(
    model=os.environ["FENNO_MODEL"],
    messages=[{"role": "user", "content": "Reply with OK only."}],
)
print(response.choices[0].message.content)
python example.py

JavaScript / TypeScript

This Node.js .mjs example runs directly. TypeScript applications use the same client configuration.

npm install openai
example.mjs
import OpenAI from 'openai'
 
const client = new OpenAI({
  baseURL: 'https://api.fenno.ai/v1',
  apiKey: process.env.FENNO_API_KEY
})
 
const response = await client.chat.completions.create({
  model: process.env.FENNO_MODEL,
  messages: [{ role: 'user', content: 'Reply with OK only.' }]
})
console.log(response.choices[0].message.content)
node example.mjs

Streaming output

Replace the request in the Python example, reusing its client:

stream = client.chat.completions.create(
    model=os.environ["FENNO_MODEL"],
    messages=[{"role": "user", "content": "Explain streaming in one sentence."}],
    stream=True,
)
for chunk in stream:
    if chunk.choices:
        text = chunk.choices[0].delta.content
        if text:
            print(text, end="", flush=True)
print()

Responses

For a model that supports Responses, reuse the client to call /v1/responses. Change the method, not the base URL; do not set it to the complete /responses endpoint:

response = client.responses.create(
    model=os.environ["FENNO_MODEL"],
    input="Reply with OK only.",
)
print(response.output_text)

Migration checklist

ItemCheck
Address fieldbase_url in Python; baseURL in JavaScript / TypeScript
Credential fieldapi_key in Python; apiKey in JavaScript / TypeScript
ModelUse an exact Fenno ID; do not invent a prefix such as openai/
CapabilitiesTools, images and structured output must be supported by the selected model
RetriesAccount for SDK retries before adding application-level retries
VerificationMatch the model, request time and usage in the Fenno console

These examples are server-side only. Do not expose an API key to run them in a browser. An SDK method does not imply Fenno supports every OpenAI hosted tool or endpoint; check the API documentation and model capabilities.

See the OpenAI-compatible API, errors and retries, and official OpenAI documentation  for SDK installation and basic usage.

ChatboxLangChain