Integrations

OpenAI SDKs

Official Python and Node SDKs: set the base URL and key. Nothing else changes. See the quickstart.

export OPENAI_BASE_URL="https://venym.io/v1"export OPENAI_API_KEY="vk-..."

Both SDKs read those two environment variables, so unmodified code can switch by environment alone.

LangChain

Use the OpenAI chat class with a custom base URL.

1from langchain_openai import ChatOpenAI2 3llm = ChatOpenAI(model="venym/qwen-27b", base_url="https://venym.io/v1", api_key="vk-...")4print(llm.invoke("Hello").content)

LlamaIndex

Use the OpenAI-compatible LLM class (OpenAILike), since stock OpenAI validates model names against OpenAI's own list.

from llama_index.llms.openai_like import OpenAILike llm = OpenAILike(model="venym/qwen-27b", api_base="https://venym.io/v1", api_key="vk-...", is_chat_model=True)

SillyTavern

• API: Chat Completion, source: Custom (OpenAI-compatible).
• Custom endpoint (base URL): https://venym.io/v1
• Custom API key: your vk- key.
• Model: enter venym/glm-flash or venym/qwen-27b, or pick from the fetched list.

For roleplay, venym/glm-flash is the cheaper, faster pick.

Open WebUI

• Admin settings, Connections, add an OpenAI API connection.
• URL: https://venym.io/v1. Key: your vk- key.
• Save, then pick the models from the model selector.

LibreChat

Add a custom endpoint in librechat.yaml:

1endpoints:2  custom:3    - name: "Venym"4      apiKey: "${VENYM_API_KEY}"5      baseURL: "https://venym.io/v1"6      models:7        default: ["venym/qwen-27b", "venym/glm-flash"]8        fetch: true9      titleConvo: true10      titleModel: "venym/glm-flash"

Editors and agents

Continue

1models:2  - name: Venym Qwen 27B3    provider: openai4    model: venym/qwen-27b5    apiBase: https://venym.io/v16    apiKey: vk-...

Cline and Roo Code

• API provider: OpenAI Compatible.
• Base URL: https://venym.io/v1. API key: your vk- key.
• Model ID: venym/qwen-27b. Set the context window to the value on the models page.

Aider

export OPENAI_API_BASE="https://venym.io/v1"export OPENAI_API_KEY="vk-..."aider --model openai/venym/qwen-27b

The openai/ prefix tells Aider to use the OpenAI-compatible route. The rest is the Venym id.

When something does not work

• Check the base URL ends in /v1, with no trailing /chat/completions.
• A 401 means the key was not sent or is wrong. A 404 on the model means the id is wrong. See errors.
• Clients that require a fixed list of OpenAI model names will not accept our ids. Use one with a custom model field.
• Menus and option names differ between client versions. If a field above has moved, look for the equivalent "base URL" and "API key" settings.