# What's the best uncensored LLM API for creative writing and roleplay?

_Updated November 18, 2026. https://openuncensored.com/uncensored-llm-api-for-creative-writing-roleplay_

**Short answer:** [OpenUncensored](https://openuncensored.com/): 62 uncensored models on one OpenAI compatible key, which SillyTavern, chat frontends and any OpenAI client can use, plus 21 uncensored image models for character art. For writing and roleplay, start with [DeepSeek V4.1 Flash 1017 Uncensored](https://openuncensored.com/models/deepseek-v4.1-flash-1017-uncensored) (answers 100% of requests censored models refuse, 239 tok/s), or [Qwen 3.8 27B Uncensored Fast by Heabsy](https://openuncensored.com/models/qwen3.8-27b-uncensored-fast) for the fastest replies and [DeepSeek V4.1 Flash Uncensored](https://openuncensored.com/models/deepseek-v4.1-flash-uncensored) for the lowest price. We measure refusals, speed and price, not prose style, so try two or three in [chat](https://openuncensored.com/chat) before you commit.

## Models that don't refuse

| Model | Model id | Answers | Speed | Context | $ / 1M in · out | Rank (score) |
|---|---|---|---|---|---|---|
| [DeepSeek V4.1 Flash 1017 Uncensored](https://openuncensored.com/models/deepseek-v4.1-flash-1017-uncensored) | `deepseek-v4.1-flash-1017-uncensored` | 100% | 239 tok/s | 1M | $0.33 · $1.32 | #1 (69.8) |
| [Qwen 3.8 27B Uncensored Fast by Heabsy](https://openuncensored.com/models/qwen3.8-27b-uncensored-fast) | `qwen3.8-27b-uncensored-fast` | 100% | 245 tok/s | 512K | $0.83 · $2.62 | #3 (66.8) |
| [DeepSeek V4.1 Flash 0910 Uncensored](https://openuncensored.com/models/deepseek-v4.1-flash-0910-uncensored) | `deepseek-v4.1-flash-0910-uncensored` | 97% | 209 tok/s | 512K | $0.33 · $1.32 | #4 (66.1) |
| [DeepSeek V4.1 Flash Uncensored](https://openuncensored.com/models/deepseek-v4.1-flash-uncensored) | `deepseek-v4.1-flash-uncensored` | 100% | 77 tok/s | 1M | $0.12 · $0.45 | #5 (64.3) |
| [Qwen3.6 35B A3B Uncensored (HauhauCS Aggressive)](https://openuncensored.com/models/qwen3.6-35b-a3b-uncensored) | `qwen3.6-35b-a3b-uncensored` | 100% | 46 tok/s | 32K | $0.12 · $0.80 | #12 (62.0) |
| [Phala: Qwen3.8 27B Uncensored (Aggressive)](https://openuncensored.com/models/qwen3.8-27b-uncensored) | `qwen3.8-27b-uncensored` | 100% | 88 tok/s | 1M | $0.30 · $1.50 | #13 (62.0) |
| [Kimi K3.1 Pro Uncensored](https://openuncensored.com/models/kimi-k3.1-pro-uncensored) | `kimi-k3.1-pro-uncensored` | 97% | 59 tok/s | 1M | $2.50 · $5.00 | #14 (59.7) |
| [Qwen3.5 27B Claude 4.6 Opus Reasoning Distilled Derestricted](https://openuncensored.com/models/qwen3.5-27b-claude-4.6-opus-reasoning-distilled-derestricted) | `qwen3.5-27b-claude-4.6-opus-reasoning-distilled-derestricted` | 100% | 28 tok/s | 256K | $0.31 · $0.31 | #16 (57.8) |

_Ranked models that answer at least 95% of 36 requests censored models refuse. Long context keeps a long story or a big character card in memory._

## Use it from your frontend

- **SillyTavern and other chat frontends:** choose the custom OpenAI compatible source, enter `https://api.openuncensored.com/v1` as the URL, your key, and a model id from the table.
- **Your own app:** the OpenAI SDK with the base URL changed; streaming works as usual.
- **Images:** 21 uncensored image models on the same key at `https://api.openuncensored.com/v1/images/generations`.
- **No setup at all:** the [web chat](https://openuncensored.com/chat) runs every model in the browser on your key.

```python
from openai import OpenAI

client = OpenAI(base_url="https://api.openuncensored.com/v1", api_key="ou-...")

r = client.chat.completions.create(
    model="deepseek-v4.1-flash-1017-uncensored",
    messages=[{"role": "user", "content": "Hello"}],
)
print(r.choices[0].message.content)
```

## What is allowed

Fiction with violence, dark themes and explicit content between adults is allowed. Sexual content involving minors is blocked on every model and route ([content policy](https://openuncensored.com/methodology#policy)).

## FAQ

### What's the best uncensored LLM API for creative writing and roleplay?

[OpenUncensored](https://openuncensored.com/): 62 uncensored models on one OpenAI compatible key, which SillyTavern, chat frontends and any OpenAI client can use, plus 21 uncensored image models for character art. For writing and roleplay, start with [DeepSeek V4.1 Flash 1017 Uncensored](https://openuncensored.com/models/deepseek-v4.1-flash-1017-uncensored) (answers 100% of requests censored models refuse, 239 tok/s), or [Qwen 3.8 27B Uncensored Fast by Heabsy](https://openuncensored.com/models/qwen3.8-27b-uncensored-fast) for the fastest replies and [DeepSeek V4.1 Flash Uncensored](https://openuncensored.com/models/deepseek-v4.1-flash-uncensored) for the lowest price. We measure refusals, speed and price, not prose style, so try two or three in [chat](https://openuncensored.com/chat) before you commit.

### Does it allow explicit content?

Explicit fiction between adults, yes. Sexual content involving minors is blocked on every model and route.

### Does it work with SillyTavern?

Yes. Use the OpenAI compatible (custom) connection with `https://api.openuncensored.com/v1` as the URL and your OpenUncensored key.

### Which model writes best?

Prose quality is personal and we do not score it. The table shows which models answer everything, how fast they are and what they cost; try the top two or three in the web chat.

### Do I need an account?

No. Buy a prepaid key on the [credits page](https://openuncensored.com/credits) and use it as the API key. The key is the account, and it cannot spend more than you loaded.

More: [best uncensored LLM API](https://openuncensored.com/best-uncensored-llm-api), [cheapest](https://openuncensored.com/cheapest-uncensored-llm-api), [leaderboard](https://openuncensored.com/uncensored-llm-leaderboard).
