# DeepSeek V4.1 Flash Abliterated

## Best for

Uncensored coding and reasoning with a long context

## When not to use

Production traffic today — it is not served yet. Workloads that rely on the model refusing unsafe requests; keep your own guardrails in front of it.

## Overview

DeepSeek V4.1 Flash Abliterated is DeepSeek V4.1 Flash with its refusal behaviour removed by abliteration, designed to answer requests the original model declines. It keeps the base model's coding, reasoning, and tool use, with a 1,048,576-token context window.

This model is not routable yet. The identifier and rates below are what it will launch with.

| | |
| --- | --- |
| API identifier | `deepseek-v4.1-flash-abliterated` |
| Developer | DeepSeek |
| Context window | 1,048,576 tokens |
| Supports | text, reasoning, tool-use, abliterated |
| Status | Not serving |

## Rates

USD per 1M tokens, on-demand tier.

| Input | Cached input | Output |
| --- | --- | --- |
| $0.24 | $0.01 | $0.96 |

## Call it

Put `deepseek-v4.1-flash-abliterated` in the `model` field. The same request works against the Anthropic Messages API — see [API Compatibility](https://inference.boundless.network/docs/api-compatibility).

```bash
curl https://api.inference.boundless.network/v1/chat/completions \
  -H "Authorization: Bearer $BOUNDLESS_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "deepseek-v4.1-flash-abliterated",
    "messages": [{
      "content": "Say hello",
      "role": "user"
    }]
  }'
```

Every model we serve is listed at [Models](https://inference.boundless.network/models); the rate card in full is on [Pricing](https://inference.boundless.network/docs/pricing).