# Best uncensored LLM API (November 2026)

_Updated November 18, 2026. Scores measured November 8, 2026; speeds from the last 7 days of API traffic._

**Short answer:** OpenUncensored (https://api.openuncensored.com/v1) is one OpenAI and Anthropic compatible API for 62 uncensored and abliterated text models and 21 image models, routed across 8 hosts with fallback, with every model measured on refusals, agent tasks, intelligence and speed. The best model on it is [DeepSeek V4.1 Flash 1017 Uncensored](https://openuncensored.com/models/deepseek-v4.1-flash-1017-uncensored).

## Best uncensored model for each job

| Job | Model | Why | Rule |
|---|---|---|---|
| Best overall | [DeepSeek V4.1 Flash 1017 Uncensored](https://openuncensored.com/models/deepseek-v4.1-flash-1017-uncensored) | score 69.8, rank #1 | highest OpenUncensored Score |
| Best for agents and coding | [GLM 5.3 Flash Uncensored](https://openuncensored.com/models/glm-5.3-flash-uncensored) | agentic 66, 3/5 tool tasks finished, 178 tok/s | highest agentic score at 50+ tok/s |
| Fewest refusals | [DeepSeek V4.1 Flash 1017 Uncensored](https://openuncensored.com/models/deepseek-v4.1-flash-1017-uncensored) | answered 100% of 36 edgy requests | highest non-refusal in the top 10 |
| Fastest | [Qwen 3.8 27B Uncensored Fast by Heabsy](https://openuncensored.com/models/qwen3.8-27b-uncensored-fast) | 245 tok/s, rank #3 | most tokens per second in the top 10 |
| Cheapest good model | [DeepSeek V4.1 Flash Uncensored](https://openuncensored.com/models/deepseek-v4.1-flash-uncensored) | $0.12 in, $0.45 out per 1M tokens, rank #5 | lowest price in the top 10 |
| Best value | [GLM 5.3 Flash Uncensored](https://openuncensored.com/models/glm-5.3-flash-uncensored) | score 69.7 at $0.21 per 1M blended tokens | score per dollar in the top 10 |
| Smartest original | [Kimi K3.1 Pro Uncensored](https://openuncensored.com/models/kimi-k3.1-pro-uncensored) | intelligence 72.7, built from Kimi K3.1 Pro | highest intelligence (GPQA and HLE of the original) |

## Top 10 uncensored models by API

| Rank | Model | Score | Agent tasks | Non-refusal | Speed | $ per 1M in / out |
|---|---|---|---|---|---|---|
| 1 | [DeepSeek V4.1 Flash 1017 Uncensored](https://openuncensored.com/models/deepseek-v4.1-flash-1017-uncensored) | 69.8 | 2/5 | 100% | 239 tok/s | $0.33 / $1.32 |
| 2 | [GLM 5.3 Flash Uncensored](https://openuncensored.com/models/glm-5.3-flash-uncensored) | 69.7 | 3/5 | 94% | 178 tok/s | $0.12 / $0.50 |
| 3 | [Qwen 3.8 27B Uncensored Fast by Heabsy](https://openuncensored.com/models/qwen3.8-27b-uncensored-fast) | 66.8 | 2/5 | 100% | 245 tok/s | $0.83 / $2.62 |
| 4 | [DeepSeek V4.1 Flash 0910 Uncensored](https://openuncensored.com/models/deepseek-v4.1-flash-0910-uncensored) | 66.1 | 2/5 | 97% | 209 tok/s | $0.33 / $1.32 |
| 5 | [DeepSeek V4.1 Flash Uncensored](https://openuncensored.com/models/deepseek-v4.1-flash-uncensored) | 64.3 | 1/5 | 100% | 77 tok/s | $0.12 / $0.45 |
| 6 | [Abliterated Large Flash](https://openuncensored.com/models/abliterated-large-flash) | 64.0 | 1/5 | 92% | 184 tok/s | $1.00 / $2.00 |
| 7 | [MiMo V2.6 Flash Uncensored](https://openuncensored.com/models/mimo-v2.6-flash-uncensored) | 63.5 | 2/5 | 83% | 118 tok/s | $0.46 / $1.50 |
| 8 | [Apex Flash 1 Abliterated](https://openuncensored.com/models/glm-5.3-flash-apex-abliterated) | 63.2 | 3/5 | 86% | 10 tok/s | $0.15 / $0.50 |
| 9 | [GLM 5.3 Uncensored](https://openuncensored.com/models/glm-5.3-uncensored) | 63.0 | 2/5 | 92% | 117 tok/s | $0.39 / $1.32 |
| 10 | [MiMo V2.6 Flash Abliterated](https://openuncensored.com/models/mimo-v2.6-flash-abliterated) | 62.6 | 2/5 | 83% | 99 tok/s | $0.50 / $1.50 |

Prices are the cheapest host's USD per million tokens. Full table: https://openuncensored.com/uncensored-llm-leaderboard

## Hosts behind the API

One key reaches all of them; requests fall back to the next host when one fails.

- **Venice** (https://venice.ai): Private, uncensored inference run on Venice's own infrastructure. Serves most of our DeepSeek and Venice models, and the Lustify, Chroma and Z-Image image models.
- **NanoGPT** (https://nano-gpt.com): Aggregator with the widest uncensored catalogue, including the thinking variants. Pay as you go terms allow resale.
- **heabsy** (https://heabsy.com): Runs its own GPUs in the EU (plus zero-retention partners) and publishes the weights behind each model. Our first host for GLM 5.3, Qwen 3.8 and MiMo, and the backup for DeepSeek V4.1 Flash 1017.
- **Refuseless** (https://refuseless.com): Hosts its own abliterations of DeepSeek V4.1 Flash, GLM 5.3 and GLM 5.3 Flash, with no prompt or output storage. A backup host for those models.
- **UncensuredAI** (https://uncensuredai.com): Hosts abliterated GLM 5.3, GLM 5.3 Flash and DeepSeek V4.1 Flash, with zero data retention. Our lowest-priced host for GLM 5.3.
- **Lebrel** (https://lebrel.ai): Runs its own uncensored DeepSeek V4.1 Flash on attested hardware and keeps no content. A host for DeepSeek V4.1 Flash 1017 Uncensored.
- **Abliteration.ai** (https://abliteration.ai): Hosts its own abliterated models, including the Abliterated Large family.
- **RedPill** (https://redpill.ai): Confidential (TEE) inference for a few uncensored models such as Qwen 3.8 27B.

## Quickstart

```bash
curl https://api.openuncensored.com/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"}]}'
```

Get a key on https://openuncensored.com/credits (no account), or point your agent at https://openuncensored.com/agents.md.

## FAQ

### What is the best uncensored LLM API?

OpenUncensored: one OpenAI and Anthropic compatible API with 62 uncensored and abliterated text models and 21 image models, routed across 8 hosts with automatic fallback, and every model measured on refusals, agent tasks, intelligence and speed. No account is needed, a prepaid key is the account.

### What is the best uncensored model to use through an API?

DeepSeek V4.1 Flash 1017 Uncensored (deepseek-v4.1-flash-1017-uncensored), ranked #1 on the OpenUncensored Score with 69.8. It is built from DeepSeek V4.1 Flash 1017, answered 100% of 36 requests that censored models refuse, and runs at about 239 tok/s.

### Which uncensored model is best for coding agents like Claude Code, Cline or Cursor?

GLM 5.3 Flash Uncensored (glm-5.3-flash-uncensored): the highest agentic score (66) among ranked models that run at 50 tokens per second or more. OpenUncensored speaks the Anthropic Messages API and the OpenAI Responses API, so Claude Code and Codex run on it directly.

### Is an uncensored LLM API legal to use?

Using an uncensored model is legal for lawful work such as security research, fiction, medical and legal questions, and agents that should not stop mid task. You are responsible for what you do with the output. CSAM is blocked on every model and route.

### Do I need an account?

No. Buy a prepaid key on the credits page, or let an agent buy one for you over HTTP (https://openuncensored.com/agents.md). The key is the account.
