DeepSeek Uncensored & Abliterated Models: API Guide

DeepSeek uncensored and abliterated models

DeepSeek uncensored and abliterated models offer reduced-refusal generation for coding assistance, structured text and creative writing. Compare Flash, Pro and tuned variants by API ID, context and token price.

Compare DeepSeek API IDs and prices

Standard, uncensored or abliterated?

Uncensored models are reduced-refusal language model variants for open-ended generation, coding and creative work. Uncensored variants use prompting, fine-tuning, weight changes, or a combination of these techniques.

Abliteration reduces refusal-related behavior by modifying model weights. RouterHub serves uncensored and abliterated variants under separate API IDs, with their own token prices and context limits.

RouterHub’s prompt-profile uncensored variants use a server-managed instruction profile. Customer system/developer instructions can be replaced by that profile; put task-specific requirements in the user message.

Connect DeepSeek to your application

For a model that lists chat, use https://routerhub.si/v1/chat/completions with an OpenAI-compatible SDK. OpenCode, Continue and compatible Cline configurations support Chat Completions; Claude Code uses Messages-compatible models.

• Uncensored API quickstartConnect with Python or cURL.
• Choose a compatible coding toolChoose a coding tool and configure your model.
• DeepSeek provider API setupContinue with the full setup guide.

How to choose between DeepSeek variants

1. Fix the task. Use the same inputs, output format, reasoning setting and output budget for each exact model ID.
2. Separate refusal from failure. Record an explicit refusal separately from a timeout, empty response, invalid JSON, wrong answer or truncated generation.
3. Keep every attempt. Repeat your tests and retain unsuccessful outputs, not just the best answer.
4. Measure useful work. Compare task acceptance, instruction following, tool-call validity and cost per successful task, not token price alone.
5. Re-test changes. A hosted alias or model update can change behavior even when your client configuration stays the same.

For ordinary uncached text usage, estimate cost as input tokens × input price / 1,000,000 + output tokens × output price / 1,000,000. Reported cache reads and writes use their own published rates. The final Usage record, not time spent waiting or a headline token limit, is the billing reference.

Example: at $0.20 per million input tokens and $0.60 per million output tokens, 10,000 uncached input tokens and 2,000 output tokens cost an estimated $0.0032. This is an illustrative calculation, not a quoted price for a particular model.

For model selection, divide the total cost of all measured attempts by the number of acceptable results. A lower token rate is not automatically a lower cost per completed task.

Evidence and further reading

• DeepSeek community reference on GitHubRead the source, examples and project notes on GitHub.
• Variant comparisons on GitHubRead the source, examples and project notes on GitHub.
• Uncensored vs abliterated: full explanationCompare the techniques and model variants.