Cursor & editors

Any editor or chat UI that accepts a custom OpenAI base URL can use api14. You need three values:

Tip: create a separate key per tool with a daily USD limit so one runaway agent cannot spend your whole balance. See Rate limits.

Cursor

1. Open Cursor Settings → Models.
2. Under API Keys, paste your sk-api14-... key into OpenAI API Key and enable it.
3. Turn on Override OpenAI Base URL and enter https://api14.com/api/v1.
4. Click Add model and enter the exact model ids you want, for example anthropic/claude-sonnet-4.5 or qwen/qwen3-coder.
5. Click Verify, then pick the model in the chat or agent model menu.

Cursor sends all OpenAI-key traffic to the override URL, so disable built-in OpenAI model names you do not want routed through api14.

Continue

Add a model to ~/.continue/config.yaml:

name: api14
version: 1.0.0
schema: v1
models:
  - name: Claude Sonnet 4.5 (api14)
    provider: openai
    model: anthropic/claude-sonnet-4.5
    apiBase: https://api14.com/api/v1
    apiKey: sk-api14-...
    roles: [chat, edit, apply]
  - name: Autocomplete (api14)
    provider: openai
    model: mistralai/codestral-2508
    apiBase: https://api14.com/api/v1
    apiKey: sk-api14-...
    roles: [autocomplete]
  - name: Embeddings (api14)
    provider: openai
    model: openai/text-embedding-3-small
    apiBase: https://api14.com/api/v1
    apiKey: sk-api14-...
    roles: [embed]

Cline

1. Open the Cline panel and click the settings gear.
2. Set API Provider to OpenAI Compatible.
3. Base URL: https://api14.com/api/v1
4. API Key: your sk-api14-... key.
5. Model ID: e.g. anthropic/claude-sonnet-4.5.
6. Optionally set the context window and enable image support to match the model's page on /models.

The same steps work for Roo Code and other Cline forks.

Open WebUI

In the UI: Admin Settings → Connections → OpenAI API, add a connection with URL https://api14.com/api/v1 and your key. Open WebUI calls GET /models, so every model on api14 appears in the model picker.

Or configure it with environment variables:

docker run -d -p 3000:8080 \
  -e OPENAI_API_BASE_URL=https://api14.com/api/v1 \
  -e OPENAI_API_KEY=sk-api14-... \
  ghcr.io/open-webui/open-webui:main

The model list is long (hundreds of models); use Admin Settings → Models to hide the ones you do not need.

Other tools

Anything built on the OpenAI SDK usually honours the standard environment variables:

export OPENAI_BASE_URL="https://api14.com/api/v1"
export OPENAI_API_KEY="sk-api14-..."

If a tool rejects model ids containing /, it is validating against OpenAI's own list, look for a "custom model" option. See Model ids & variants.