Everything you need.

Quick start

1. 1

   Create an account

   Sign up and confirm your email.

2. 2

   Add credit

   From 2 €, no subscription. New accounts start at 0.

3. 3

   Create an API key

   Under Developers. The key (orv_…) is shown only once.

4. 4

   Get going

   Via the API, the CLI or tools like opencode.

API

The API is OpenAI compatible. Any library that speaks OpenAI works with a changed base URL.

Base URL

https://orrevan.com/api/v1

Authentication

Header Authorization: Bearer orv_… on every request. Never use the key in a browser or in public code.

Endpoints

GET /models

All models with prices and capabilities. No key needed.

POST /chat/completions

Chat answer, optionally streamed.

GET /me

Account and credit for the key.

POST /web/search

Web search, billed per search with results.

curl

curl https://orrevan.com/api/v1/chat/completions \
  -H "Authorization: Bearer $ORREVAN_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model": "orrevan/abliterated-model",
       "messages": [{"role": "user", "content": "Hi"}]}'

from openai import OpenAI

client = OpenAI(base_url="https://orrevan.com/api/v1", api_key="orv_...")
r = client.chat.completions.create(
    model="orrevan/abliterated-model",
    messages=[{"role": "user", "content": "Hi"}],
)
print(r.choices[0].message.content)

import OpenAI from "openai";

const client = new OpenAI({ baseURL: "https://orrevan.com/api/v1", apiKey: process.env.ORREVAN_API_KEY });
const r = await client.chat.completions.create({
  model: "orrevan/abliterated-model",
  messages: [{ role: "user", content: "Hi" }],
});
console.log(r.choices[0].message.content);

Models

Model IDs start with orrevan/. Prices in USD per 1M tokens, live from the system.

Every model is uncensored. Images as base64 data URL (PNG, JPEG, WebP) in the image_url field.

Streaming

With "stream": true the answer arrives as server-sent events, piece by piece. With stream_options.include_usage you get the token counts at the end.

stream = client.chat.completions.create(
    model="orrevan/abliterated-model",
    messages=[{"role": "user", "content": "Hi"}],
    stream=True,
    stream_options={"include_usage": True},
)
for chunk in stream:
    if chunk.choices and chunk.choices[0].delta.content:
        print(chunk.choices[0].delta.content, end="")

Tools and extras

tools, tool_choice

Function calling in OpenAI format, up to 128 tools per request.

reasoning_effort

Thinking depth: none, low, medium, high or max, depending on the model (see /models, capabilities.reasoning).

response_format

JSON output where the model supports it.

web_search: true

Orrevan extra: searches the web before answering. Small fee per search.

temperature, top_p, max_tokens, stop, seed

Same as OpenAI.

{"model": "orrevan/abliterated-model",
 "messages": [{"role": "user", "content": "Weather in Tallinn?"}],
 "tools": [{"type": "function", "function": {
   "name": "get_weather",
   "parameters": {"type": "object",
     "properties": {"city": {"type": "string"}}, "required": ["city"]}}}],
 "reasoning_effort": "low"}

Error codes

400

Invalid request, e.g. broken JSON or unknown model.

401

Key missing, wrong, revoked or expired.

402

Credit used up. Top up under Credit.

413

Request too large.

429

Too many requests, too many in parallel or the key's daily limit is reached.

502, 503, 504

Model not reachable right now. Retry later, nothing is charged.

CLI

Uncensored coding agent for your terminal. Needs Node.js 20 or newer.

npm i -g orrevan
orrevan login
orrevan

opencode

opencode works with Orrevan as a provider. Create ~/.config/opencode/opencode.json, set the key as an environment variable, start opencode.

Set the key (e.g. in ~/.bashrc or ~/.zshrc)

export ORREVAN_API_KEY="orv_..."

Configuration

{
  "$schema": "https://opencode.ai/config.json",
  "provider": {
    "orrevan": {
      "npm": "@ai-sdk/openai-compatible",
      "name": "Orrevan",
      "options": {
        "baseURL": "https://orrevan.com/api/v1",
        "apiKey": "{env:ORREVAN_API_KEY}"
      },
      "models": {
        "orrevan/abliterated-model": { "name": "Abliterated Model" }
      }
    }
  },
  "model": "orrevan/orrevan/abliterated-model"
}

Add more models to the models block. GET /models lists every ID.

Limits and key expiry

60 requests per minute

Per key, can be set lower.

3 parallel requests

Per account, across all keys.

Daily limit

Default $50 per key and day ($0.50 to $1,000). Resets at midnight UTC.

Expiry

Every key expires, after 1, 3, 6 or 12 months. No "never". Email reminder 7 days before.

Renewal

Rotating a key creates a new one with the same limits, the old one stops working.

Alerts

Email on unusual spending, new IPs and a reached daily limit.

Pricing and credit

Pay as you go

You pay per token. No subscription, no base fee.

Top up

2, 5, 10 or 25, or any amount from 2 to 500, in EUR or USD.

One credit

Web chat, API and CLI share the same credit.

Errors are free

If a request fails, nothing is charged.

Privacy

• No training on your data. Orrevan does not use your prompts or answers to train AI models.
• API requests are not stored. Only billing data is kept (model, tokens, cost).
• Web chats are stored encrypted with AES-256-GCM and can be deleted at any time.
• The CLI keeps conversations on your machine only.