Skip to content
Bkper
Sign In

Overview

Bkper AI API (v1)

Public inference API implementing the stateless Bkper Open Responses 2026-04-24 profile.

Base URL

https://ai.bkper.app

OpenAPI specification

The canonical machine-readable contract is available at https://ai.bkper.app/openapi.json.

Authentication

Send a Bkper OAuth access token with every inference request:

Authorization: Bearer <Bkper access token>

See the Bkper AI provider guide for token setup and client configuration.

Request workflow

  1. Call GET /v1/models to discover the current public model IDs, capabilities, limits, and usage rates.
  2. Select a returned model ID.
  3. Call POST /v1/responses with that model and explicit input.

Use stream: false for a complete JSON response. Use stream: true for semantic server-sent events ending with data: [DONE].

Stateless operation and privacy

Bkper AI does not persist response state. Omitted store defaults to false; the only accepted explicit value is store: false. Continue a conversation by sending explicit prior items in input.

Bkper usage logs record attribution, status, token, cache, and cost metadata. They exclude prompt and response content. Provider-specific retention and caching boundaries are documented in the Bkper AI provider guide.

Usage

Requests consume the authenticated user’s Bkper AI allowance. The model catalog provides current effective rates and limits. New requests stop when the recorded allowance is exhausted; Bkper does not create automatic paid AI overages. See Models and Usage for policy details.

Attribution

Clients may send the optional bkper-ai-source header with a stable lowercase application identifier so usage can be attributed to that client.

Bkper AI provider documentation

  • OpenAPI version: 3.1.0

Authentication

bkperBearer

Send a Bkper OAuth access token. See token setup.

Security scheme type: http

Bearer format: Bkper OAuth access token