Tool
Uncensored LLM API price comparison
Describe one typical request and your monthly volume. Every listed offering is priced with the same formula, and the table says which rows serve the same weights, so you can tell a real saving from a different model.
Disclosure. Uncensored AI Atlas has a commercial relationship with MolochAI. We may benefit when you sign up or buy credit through our links. MolochAI rows are labelled Partner and sorted by cost like every other row. How we handle this.
| Provider and model id | What is served | In / Out / Cached per 1M | Per request | Per month | Context | Checked |
|---|---|---|---|---|---|---|
| MolochAIPartner qwen3.8-27b-uncensored | Own refusal-reduced build, weights not named | $0.50 / $2.00 / not listed | $0.0027 | $81.00 | 200,704 | 14 November 2026 |
| Featherless huihui-ai/Huihui-Qwen3.8-27B-abliterated | Names the exact Hugging Face repo huihui-ai/Huihui-Qwen3.8-27B-abliterated | $1.60 / $12.00 / not listed | $0.0120 | $360.00 | 32,768 | 14 November 2026 |
When rows are comparable
Two rows are a like-for-like price comparison only when they serve the same weights. The table marks rows that name the same Hugging Face repo. A provider's own build may be a different variant, a different quantization, or both; a standard model is a different set of weights altogether. Context limits also differ between providers for the same weights, which can matter more than price for long prompts.
The CSV contains exactly the rows shown, with the rates, your workload, the computed costs, the date each rate was checked and its source URL, so a comparison can be checked later. Methodology for the rates: price table. For a single offering in more detail, use the token cost calculator.