API

Proxify API for hosted agents and OpenCode

Call four concrete models through Chat Completions. Every model supports native client tool calls for OpenCode and other compatible clients.

Base URL https://ai.proxify.host/api

Agent from 1 token / request

proxify-agent-1.0

Fast Proxify agent for chat, code, analysis, tools, and everyday integrations.

Pro Agent from 3 tokens / request

proxify-superagent-1.0

Advanced Proxify agent for complex reasoning, coding, research, and long-running server tasks.

How it works

Create an account and open the API section in settings.
Generate a bearer key once and save it securely.
Call https://ai.proxify.host/api/v1/models and https://ai.proxify.host/api/v1/chat/completions.
Requests are billed from a separate API balance in RUB inside your account.

API balance top-ups, support and custom volume are handled via @proxifyaihelp.

Choose any model returned by GET /v1/models for OpenCode and other OpenAI-compatible coding clients with native tool calls. Hosted integrations and connectors remain unavailable through the API.

Quick start

curl https://ai.proxify.host/api/v1/chat/completions \
  -H "Authorization: Bearer pxy_live_xxxxxxxxxxxxx" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "openai/gpt-5.6-sol",
    "messages": [
      {"role": "user", "content": "Собери короткий план запуска AI-продукта."}
    ],
    "web_search": false
  }'

Proxify API for hosted agents and OpenCode

Call four concrete models through Chat Completions. Every model supports native client tool calls for OpenCode and other compatible clients.

Open Proxify AI Message on Telegram