Connect to Webparam

An OpenAI-compatible endpoint. If your tool speaks the OpenAI API, it already speaks Webparam.

The two values you need

Base URL: https://webparam.org/v1
API key: wpk_live_…
Header: Authorization: Bearer <key>

Keep the key in an environment variable. Webparam stores a hash of it and shows the value once, at creation.

SDKs

Examples use ac-flash-4-5-standard. Any model id from your catalogue works.

# pip install openai
import os
from openai import OpenAI

client = OpenAI(
    api_key=os.environ["WEBPARAM_API_KEY"],
    base_url="https://webparam.org/v1",
)

response = client.chat.completions.create(
    model="ac-flash-4-5-standard",
    messages=[{"role": "user", "content": "Hello"}],
)

print(response.choices[0].message.content)

Editors and agents

Continue, VS Code and anything else that accepts an OpenAI-compatible endpoint.

# ~/.continue/config.yaml
name: Webparam
version: 1.0.0
models:
  - name: ac-flash-4-5-standard
    provider: openai
    model: ac-flash-4-5-standard
    apiBase: https://webparam.org/v1
    apiKey: wpk_live_your_key_here
    roles:
      - chat
      - edit
      - apply

Model ids come from GET https://webparam.org/v1/models, which returns only the models your key is allowed to call.

What the API supports

Chat completions: POST /chat/completions
Streaming: server-sent events
Embeddings: POST /embeddings
Model list: GET /models

Tool calling: Passed through untouched
JSON mode: Passed through untouched
Rate limits: Per project, set by you
Billing: Per token, prepaid, never a surprise

Errors

Stable codes, and a request id on every response.

Code	HTTP	What to do
INVALID_API_KEY	401	Check the key, and that it has not been revoked.
MODEL_NOT_ALLOWED	403	The project or key policy excludes this model.
INSUFFICIENT_BALANCE	402	Add credit.
PROJECT_BUDGET_EXCEEDED	402	Raise the project budget, or wait for the reset.
KEY_SPEND_LIMIT_EXCEEDED	402	Raise or remove this key's monthly limit.
RATE_LIMITED	429	Back off; `retry-after` says how long.
MODEL_UNAVAILABLE	503	Transient. Retry, or route to another model.
UPSTREAM_TIMEOUT	504	Retry with backoff.

Every response carries x-request-id. Quote it if you contact support — it is how a failure is traced.

Create a workspace, add credit and mint a key — usually under two minutes.

Get started Read the docs