# Uncensored LLM API pricing compared

Per-million-token prices for uncensored and abliterated LLM APIs (Venice, NanoGPT, decuff, abliteration.ai), checked 8 Nov 2026, plus subscriptions, platform fees, free tiers and where we cost more.

## Short answer

Uncensored and abliterated APIs cost roughly **$0.16 to $3 per million input tokens and $0.50 to $5 per million output tokens** (8 Nov 2026). The cheapest per token are Venice’s API and NanoGPT’s uncensored routes. Unrestricted AI costs more per token ($1 to $5.44 input). We bundle chat and API into one monthly allowance across three providers.




## Per-token rates (USD per 1M tokens, checked 8 Nov 2026)





| Provider | Model | Input | Output | Context |
| --- | --- | --- | --- | --- |
| Venice API | venice-uncensored-1-2 | $0.20 | $0.90 | 128K |
| Venice API | venice-uncensored-role-play | $0.50 | $2.00 | 128K |
| Venice API | gemma-4-uncensored | $0.1625 | $0.50 | 256K |
| Venice API | deepseek-v4-1-flash-uncensored | $0.33 | $1.32 | 1000K |
| NanoGPT | GLM 5.3 Flash Uncensored | $0.20 | $0.80 | Not listed |
| NanoGPT | DeepSeek V4.1 Flash 1017 Uncensored | $0.33 | $1.32 | Not listed |
| NanoGPT | GLM 5.3 Uncensored | $1.25 | $2.25 | Not listed |
| decuff | decuffed-qwen3.8-27b (Flash) | $0.50 | $2.00 | 64K |
| decuff | decuffed-deepseek-v4.1-flash (Flagship) | $0.33 | $1.32 | 256K |
| abliteration.ai | abliterated-model | $1.00 | $3.00 | 256K |
| abliteration.ai | abliterated-model-large-flash | $1.00 | $2.00 | 1M |
| abliteration.ai | abliterated-model-large-v2 | $3.00 | $5.00 | 1M |
| **Unrestricted AI** | venice-uncensored-1-2 | $1.00 | $2.00 | 32,768 |
| **Unrestricted AI** | venice-uncensored-role-play | $1.00 | $4.00 | 32,768 |
| **Unrestricted AI** | featherless-llama-3.3-70b-instruct-abliterated | $1.30 | $2.00 | 32,768 |
| **Unrestricted AI** | featherless-qwen2.5-32b-instruct-abliterated | $5.44 | $9.60 | 32,768 |

## Plans, fees and free tiers (checked 8 Nov 2026)





| Provider | Free | Subscription | Fees and notes |
| --- | --- | --- | --- |
| Venice | 10 text prompts a day (app); API on credits | Pro $18/mo (as displayed); Pro Plus $68/mo; Max $200/mo | 100 credits = $1 |
| NanoGPT | One free model on the web | Pro $12/mo: 60M input tokens a week on included open-source models | "API at list prices — no markup"; pinned providers add about 5% |
| decuff | About 1M tokens with a phone sign-in | Developer $20/mo | Prepaid minimum $10; all payments final |
| abliteration.ai | One-credit free preview | Developer $20, Growth $50, Scale $200 per month | Cached input at 10% of the input rate |
| OpenRouter | 25+ free models | None | 5.5% platform fee on Standard |
| Featherless | None | Chat $25/mo unlimited tokens at 32K context (interactive use only); Developer from $50/mo |  |
| **Unrestricted AI** | 50 chat messages a day (chat app); one-time $0.50 API trial for verified accounts | Starter $19, Pro $100, Max $200, Ultra $500 per month, each including the same amount of usage | Top-ups from $5; no automatic overage; per-key budgets |

## Worked example

A role-play turn with 8,000 input and 300 output tokens:

- Venice API `venice-uncensored-role-play`: **$0.0046**
- Unrestricted AI, same model: **$0.0092**
- decuff Flagship: **$0.0030**


## When we are worth the premium

When we are not: if you only need the API for one Venice model, Venice’s own API is cheaper and has a larger context window.

- One allowance covers both a [chat app](/unrestricted-ai-chat) and an API key.
- Spending caps per key.
- Models from three upstream providers behind one key and one [model list](/ai-models).
- Works in [SillyTavern](/learn/sillytavern-custom-openai-endpoint) and [JanitorAI](/learn/janitorai-proxy-setup) without changing sampler settings.


## FAQ

**What is the cheapest uncensored LLM API?**

By listed rates on 8 Nov 2026, Venice’s `gemma-4-uncensored` ($0.1625 in, $0.50 out) and NanoGPT’s GLM 5.3 Flash Uncensored ($0.20, $0.80).

**Why is Unrestricted AI more expensive than Venice for the same model?**

We resell Venice routes along with NanoGPT and Featherless routes, with our own chat app, billing and key controls on top.

**Do output tokens cost more?**

On every service listed, yes.

**Are there free uncensored APIs?**

OpenRouter lists 25+ free models (not all of them uncensored), decuff gives about 1M tokens with a phone sign-in, and abliteration.ai has a one-credit preview. We give verified accounts a one-time $0.50 API trial; our 50 free daily messages cover the chat app.

## Sources

- [Venice API pricing (checked 8 Nov 2026)](https://docs.venice.ai/overview/pricing)
- [Venice pricing (checked 8 Nov 2026)](https://venice.ai/pricing)
- [NanoGPT pricing (checked 8 Nov 2026)](https://nano-gpt.com/pricing)
- [NanoGPT llms.txt (checked 8 Nov 2026)](https://nano-gpt.com/llms.txt)
- [decuff pricing (checked 8 Nov 2026)](https://decuff.ai/pricing)
- [abliteration.ai pricing (checked 8 Nov 2026)](https://abliteration.ai/pricing)
- [OpenRouter pricing (checked 8 Nov 2026)](https://openrouter.ai/pricing)
- [Featherless pricing (checked 8 Nov 2026)](https://featherless.ai/pricing)
- [Unrestricted AI model catalog (JSON)](https://tryunrestrictedai.com/api/catalog)
- [Unrestricted AI pricing](https://tryunrestrictedai.com/api/pricing)

Canonical: https://tryunrestrictedai.com/compare/uncensored-llm-api-pricing