Ropewalk API & MCP

MCP server & API for every AI model — at half the price

Text, image, video, audio and 3D models from one key, billed per use from your Ropewalk wallet. API prices are 50% of in-app prices.

Power up your AI agent

Claude, ChatGPT or Cursor make images, video and music right in the chat. One link, no code.

Connect your agent

Build with the API

OpenAI-compatible endpoints for every model, file uploads and job polling.

Quick start

Request builder

Pick a model, set its options and files, and copy a ready request.

Open the builder

50% of in-app prices

Every model costs half as much through the API as in the app. No subscription required — pay per request in coins.

All modalities

Chat models, image, video, voice, music and 3D generation behind one key and one bill.

Pay only for results

Unused token reservations come back and failed generations are refunded automatically.

One wallet

The same coins as in the app: top up by card or crypto, see every API charge in your history.

Works with
  • Claude Code
  • Claude Desktop
  • Cursor
  • Windsurf
  • VS Code
  • OpenAI SDK
  • LangChain
  • n8n

Same models, half the price

Starting prices in coins (1 USD of top-up = 1000 coins). Settings like resolution or duration can raise the price.

ModelIn the appVia API
Inkling65+33+
DeepSeek V4.1 Flash10+5+
Seedance 2.0 Mini640+320+
FLUX 31200+600+
FLUX 3 Image85+43+
Ideogram 4.5120+60+
Tripo P1 Image to 3D1600+800+
Tripo P1 Text to 3D1600+800+
MiniMax Speech 2.8 HD400+200+
MiniMax Speech 2.8 Turbo245+123+

All models and exact prices: GET /v1/models

Start in a minute

Use the OpenAI SDK you already have, plain HTTP, or connect an MCP client.

Python
from openai import OpenAI

client = OpenAI(base_url="http://ropewalk.ai/api/v1", api_key="rw_live_YOUR_KEY")

reply = client.chat.completions.create(
    model="thinkingmachines/inkling",
    messages=[{"role": "user", "content": "Write a tagline for a coffee shop"}],
)
print(reply.choices[0].message.content)
print(reply.usage)  # includes cost_coins

Questions

What is an MCP server?

MCP (Model Context Protocol) is an open standard that lets AI assistants such as Claude, Cursor or VS Code use external tools. The Ropewalk MCP server gives your assistant tools to list AI models, chat with GPT or Claude, generate images and video with FLUX, Nano Banana or Kling, and check your balance — all through one connection.

Which AI models can I use through MCP and the API?

The same catalog as the app: chat models such as GPT and Claude, image models such as FLUX and Nano Banana, video models such as Kling and Seedance, voice and music models such as ElevenLabs, and 3D models. One key and one MCP server cover all of them.

How much does the API cost?

Each request costs 50% of what the same request costs in the Ropewalk app, in coins from your wallet. Text models are charged per token actually used; media models per generation, depending on settings.

Do I need a subscription?

No. Top up your wallet with any amount and pay per request. Subscription coins can be spent through the API too.

Which models are available?

Almost every model from the Ropewalk catalog. GET /v1/models returns the current list with API prices, settings and accepted inputs.

What happens if a generation fails?

The reserved coins are refunded in full. For text models, the unused part of the token reservation is refunded as soon as the answer arrives.

Is my key safe?

Keys are stored only as a hash and shown once. Revoke a key at any time in Settings → API; it stops working immediately. Keep keys on your server, never in browser code.

Are there limits?

120 requests per minute and 5 simultaneous media generations per key. Contact us if you need more.

Build with every model today

Create a key in seconds — it works with your existing Ropewalk balance.