# Use Uncensored Models on OpenCode, Codex, and Claude Code

> Source: <https://openuncensored.com/blog/use-uncensored-models-in-cursor-cline-codex>
> Published: 2026-11-12 01:05:38+00:00

OpenUncensored is an OpenAI compatible API, so any tool that lets you set a custom OpenAI base URL can use its 82 uncensored models (60 text and 22 image). You need two things:

- **Base URL:**`https://openuncensored.com/api/v1`
- **API key:** an`ou-` key from[openuncensored.com/credits](https://openuncensored.com/credits) . No account, no sign up.

Pick a model id from the [rankings](https://openuncensored.com/rankings) or `GET https://openuncensored.com/api/v1/models`. For agent and coding work,
start with `deepseek-v4.1-flash-1017-uncensored`: it is at the top of the rankings, fast, and cheap on long agent sessions because most of
each request is served from the prompt cache.

## Let your agent set it up

The fastest way: paste this into Claude Code, Cursor, Codex, OpenCode or any agent:

Read [https://openuncensored.com/agents.md](https://openuncensored.com/agents.md) and set me up with an OpenUncensored key.

The guide tells the agent how to start a purchase, hand you the payment link, and collect the key itself.

## Cline, Kilo Code and Roo Code

In the extension's settings, choose the **OpenAI Compatible** provider, then set:

- Base URL: `https://openuncensored.com/api/v1`
- API key: your `ou-` key
- Model ID: for example `deepseek-v4.1-flash-1017-uncensored`

## Cursor

In Cursor's model settings, enter your `ou-` key as the OpenAI API key, turn on the option to override the OpenAI
base URL and set it to `https://openuncensored.com/api/v1`, then add the model id (for example `deepseek-v4.1-flash-1017-uncensored`) as a custom model.

## Codex CLI

Add a provider to `~/.codex/config.toml`:

```
[model_providers.openuncensored]
name = "OpenUncensored"
base_url = "https://openuncensored.com/api/v1"
env_key = "OU_API_KEY"
wire_api = "responses"

[profiles.uncensored]
model_provider = "openuncensored"
model = "deepseek-v4.1-flash-1017-uncensored"
```

Then `export OU_API_KEY=ou-...` and run `codex --profile uncensored`. Codex speaks the OpenAI Responses API, which
OpenUncensored serves at `/api/v1/responses`, so file edits, shell commands and multi-step tasks work as usual.

## OpenCode

Add OpenUncensored as a provider in `opencode.json`, either in your project or in `~/.config/opencode/opencode.json`
for every project:

```
{
  "$schema": "https://opencode.ai/config.json",
  "providers": {
    "openuncensored": {
      "name": "OpenUncensored",
      "env": ["OPENUNCENSORED_API_KEY"],
      "package": "@opencode/ai/providers/openai-compatible",
      "settings": { "baseURL": "https://openuncensored.com/api/v1" },
      "models": {
        "deepseek-v4.1-flash-1017-uncensored": { "name": "DeepSeek V4.1 Flash 1017 Uncensored" }
      }
    }
  }
}
```

Add any other model id from the [rankings](https://openuncensored.com/rankings) to `models` the same way. Then
`export OPENUNCENSORED_API_KEY=ou-...` and pick the model with `/models` in OpenCode, or run it directly:

```
opencode run -m openuncensored/deepseek-v4.1-flash-1017-uncensored "Explain how SQL injection works"
```

OpenCode 2 runs a background service that only sees the environment variables it started with. If OpenCode says the
key is missing after you exported it, add `--standalone` to start a private server that reads your current shell.

## Claude Code

Claude Code runs on Anthropic's models, but it can compare and call uncensored models through our MCP server:

```
claude mcp add --transport http open-uncensored https://openuncensored.com/mcp --header "Authorization: Bearer $OU_API_KEY"
```

That gives it tools to list models, read the rankings, check your balance and send a prompt to any model.

## OpenAI SDK, LangChain, LiteLLM

``` python
import os
from openai import OpenAI

client = OpenAI(base_url="https://openuncensored.com/api/v1", api_key=os.environ["OU_API_KEY"])
reply = client.chat.completions.create(
    model="deepseek-v4.1-flash-1017-uncensored",
    messages=[{"role": "user", "content": "Explain how SQL injection works."}],
)
print(reply.choices[0].message.content)
```

LangChain's `ChatOpenAI` and LiteLLM's `openai/` provider take the same base URL and key.

## Good to know

- Requests are billed per token at exactly the provider's price, with no markup; cached prompt tokens at the provider's cached price.
- If a model's providers are down, the API falls back to the next best ranked model built on the same original.
The `X-OU-Model` response header says which model answered; send`"routing": {"allow_fallbacks": false}` to
turn that off.
- Streaming, tools and the usual OpenAI parameters pass through.

## Try any of them in a minute

No account and no sign up. Buy a prepaid key at [openuncensored.com/credits](https://openuncensored.com/credits), or hand your
agent [openuncensored.com/agents.md](https://openuncensored.com/agents.md) and let it get one for you. Then:

```
curl https://openuncensored.com/api/v1/chat/completions \
  -H "Authorization: Bearer $OU_API_KEY" -H "Content-Type: application/json" \
  -d '{"model": "deepseek-v4.1-flash-1017-uncensored", "messages": [{"role": "user", "content": "Hello"}]}'
```

The full, always current table is on the [rankings page](https://openuncensored.com/rankings).