DeepSeek V4.1 Flash Uncensored
DeepSeek V4.1 Flash Uncensored is dealignai's build of DeepSeek-V4.1-Flash (MIT licence) with refusal behaviour removed directly in the weights — no fine-tuning and no adapter. Underneath is DeepSeek's September 2026 mixture-of-experts model with 552B parameters, about 8B active while reading and 16B while writing, made for a context of up to a million tokens; reasoning comes back separately from the answer, and tool calling works for agents. It also reads images, up to 8 per request. Anything in front of end users needs your own moderation. We serve it on hardware we operate outside the EEA; prompts and completions are not stored.
model id: deepseek-v4.1-flash-unc
from openai import OpenAI client = OpenAI( api_key="hb-...", base_url="https://api.heabsy.com/v1",) r = client.chat.completions.create( model="deepseek-v4.1-flash-unc", messages=[{"role": "user", "content": "Hello"}],)print(r.choices[0].message.content)
curl https://api.heabsy.com/v1/chat/completions \ -H "Authorization: Bearer $HEABSY_API_KEY" \ -H "Content-Type: application/json" \ -d '{ "model": "deepseek-v4.1-flash-unc", "messages": [{"role": "user", "content": "Hello"}] }'
import OpenAI from "openai"; const client = new OpenAI({ apiKey: process.env.HEABSY_API_KEY, baseURL: "https://api.heabsy.com/v1",}); const r = await client.chat.completions.create({ model: "deepseek-v4.1-flash-unc", messages: [{ role: "user", content: "Hello" }],});console.log(r.choices[0].message.content);
export ANTHROPIC_BASE_URL=https://api.heabsy.comexport ANTHROPIC_AUTH_TOKEN="$HEABSY_API_KEY"export ANTHROPIC_MODEL=deepseek-v4.1-flash-uncexport ANTHROPIC_DEFAULT_OPUS_MODEL=deepseek-v4.1-flash-uncexport ANTHROPIC_DEFAULT_SONNET_MODEL=deepseek-v4.1-flash-uncexport ANTHROPIC_DEFAULT_HAIKU_MODEL=deepseek-v4.1-flash-uncexport CLAUDE_CODE_MAX_CONTEXT_TOKENS=1048576claude
Use it in Claude Code · OpenCode · Cline · Kilo Code · SillyTavern · n8n
Price list, updated 29 October 2026. No minimum spend, no subscription.
About the model
- Developer
- dealignai
- Released
- September 2026
- Licence
- MIT
- Architecture
- mixture of experts · 552B parameters, 8–16B active per token
- Based on
- DeepSeek-V4.1-Flash
- Input
- text, images
- Languages named by the developer
- none named on the model card
Source: the developer's model card, as of 17 September 2026.
DeepSeek V4.1 Flash Uncensored and the 1017 build
Both are dealignai's versions of DeepSeek-V4.1-Flash, a 552B mixture-of-experts model (8B active while reading, 16B while writing) with refusals removed in the weights; the 1017 build is the October 2026 release. Both read images, up to 8 per request, and take a context of up to a million tokens.
We run both on our own GPUs with zero data retention: prompts and completions are not written to logs or durable storage, and nothing is used for training. The GPUs are outside the EEA. Anything in front of end users needs your own moderation.
Licence: what it means for your company
MIT. A permissive open-source licence: you may use, modify and sell products built on the model, including commercially; when you redistribute the weights, the licence and copyright notices go with them.
A summary of the licence, not legal advice.
Technical details
- Context window
- 1M tokens
- Max output
- 65,536 tokens
- Throughput limit
- 2M tok/min
- Tokenizer
- Other
- Streaming
- yes
- Open weights
- —
- max_tokens
- 1 … 65536
- temperature
- 0 … 2
- top_p
- 0 … 1
- frequency_penalty
- -2 … 2
- presence_penalty
- -2 … 2
- stop
- array
- seed
- integer
This model runs on GPUs we operate outside the European Economic Area. Prompts and completions are not written to our logs or durable storage, and nothing is used for training. For processing strictly inside the EEA, pick a model marked EEA. How we serve our own models is described on the inference page.
Questions about this model
›Where is DeepSeek V4.1 Flash Uncensored computed?
Outside the EEA, on GPUs we operate ourselves. Prompts and completions are not stored; if you need processing strictly inside the EEA, choose a model marked EEA in the catalog.
›Does DeepSeek V4.1 Flash Uncensored store prompts and completions?
No. This model runs with zero data retention: request content is not written to durable storage and nothing is used for training.
›How large is the DeepSeek V4.1 Flash Uncensored context window?
1M tokens, with up to 65,536 tokens of output.
›What does DeepSeek V4.1 Flash Uncensored cost?
$0.40 per million input tokens and $1.40 per million output tokens. No minimum spend and no subscription; the same price list covers the API and the invoice.
›Is there a throughput limit on DeepSeek V4.1 Flash Uncensored?
Yes, 2M tokens per minute. Higher limits are a question of volume, not of principle.
›Which languages does DeepSeek V4.1 Flash Uncensored support?
The developer's model card names no languages. Test the model on your own material before relying on it in a particular language.
›Under which licence is DeepSeek V4.1 Flash Uncensored released?
MIT. A permissive open-source licence: you may use, modify and sell products built on the model, including commercially; when you redistribute the weights, the licence and copyright notices go with them.
Compare with similar models
| Model | Developer | Parameters | Context | Licence | Input / 1M |
|---|---|---|---|---|---|
| DeepSeek V4.1 Flash Uncensored | dealignai | 552B / 8–16 | 1M | MIT | $0.40 |
| DeepSeek V4.1 Flash 1017 Uncensored | dealignai | 552B / 8–16 | 1M | MIT | $0.40 |
| DeepSeek V4.1 Flash | DeepSeek | 552B / 8–16 | 1M | MIT | $0.28 |
| DeepSeek V4 Flash | DeepSeek | 284B / 13B | 1M | MIT | $0.28 |
36 more models behind the same key and the same invoice.
Browse the catalog