How to Use DeepSeek V4.1 Flash Uncensored in LU: Chat, Agent and Code, Step by Step

DeepSeek V4.1 Flash Uncensored runs in LU Cloud, so you need no graphics card and no download. This guide shows how to reach it in the Locally Uncensored desktop app, in the browser and on your phone, and how to set up Chat, Agent and Code for it. What the model is and what its makers measured is in DeepSeek V4.1 Flash Uncensored in LU Cloud.

What You Need

• Locally Uncensored 3.0.7 on Windows, Mac or Linux, or any browser, or the LU app on Android
• An LU Cloud account with credits
• For Agent and Code: a folder for the agent to work in

Step 1: Sign In to LU Cloud

In the desktop app, open Settings, General, LU Cloud Account and click "Create account", or "Sign in" if you have one. Your browser opens on lu-labs.ai. After the login the page says "Signed in" and the app picks up the session on its own. The account card then shows your plan and your credits; "View plans" opens the pricing page if you have none.

Step 2: Switch to Cloud

Click the Cloud switch in the header next to the model picker. It turns into "Switch to Cloud?"; click again within six seconds. The picker now lists the hosted models and the composer shows a "Cloud" chip. Your local models are freed from memory and come back when you switch to Local.

Step 3: Pick the Model

Open the model picker and scroll to the Uncensored group, marked Beta. Pick DeepSeek V4.1 Flash Uncensored. The desktop app reads the hosted list from the server, so it is there without an app update.

Next to it you find Abliterated Large Flash, the other model that arrived the same day. Pick that one instead if you need to attach images: the DeepSeek build is text only in LU Cloud.

Step 4: Chat

Type as usual. Two controls next to the message box matter:

• Think. The model reasons by default. Switch Think off for short questions, rewrites and role play turns; on this model off really stops the reasoning.
• Effort. Low, Medium and High. High is the default. Thinking tokens are output tokens and output costs four times as much as input here.

Long conversations are no problem for the window, which holds one million tokens. Each turn still carries the history, so a very long chat costs more per message than a fresh one.

Step 5: Agent Mode

The model calls tools natively, so it can drive the agent.

1. Open a new, empty chat and click the "Agent" toggle above the message box before the first message.
2. In "Where should the agent work?", choose "Sandbox" for a first try, or "Pick a folder…" for a real project.
3. Describe the task with the result you want, for example: "Search the web for the three latest reviews of this game, read them and write a summary to notes.md."
4. Approve the tool calls as they come. You see which tool the model wants to call and with what.

The workspace protection is a folder jail, not a virtual machine: commands run on your computer. A model that does not refuse also does not hesitate, so read each tool request before you allow it.

Step 6: The Code Tab

1. Open the Code tab and pick your project folder with "Select folder...".
2. Leave the mode on "Ask permissions" at first: commands need your ok, and edits come as a diff before they land.
3. Pick DeepSeek V4.1 Flash Uncensored and give a coding instruction, such as "Write a script that renames the photos in this folder by the date they were taken."
4. Follow the plan bar, review each diff, then accept or reject it.

In the Browser

1. Open lu-labs.ai and sign in with the same account.
2. Open the Studio and the model picker. In the browser the group is called Unrestricted and the model DeepSeek V4.1 Flash Dealigned, after the team that made it.
3. Chat, Agent and Code work there with the same credits.

On Your Phone

• Android: install the LU app and sign in with your LU Cloud account. The model is among the hosted chat models.
• iPhone: the app is in review at the App Store. Until then, open lu-labs.ai in Safari and sign in.

What It Costs

The meter follows the live list price, so the figures can move a little. A 15 EUR pack carries 450,000 credits. Current plan and pack prices are on the pricing page.

When It Does Not Work

• The model is not in the picker. Check that the composer shows the "Cloud" chip. In Local mode the picker lists only your own models.
• A request is refused with a sentence about the route. The model runs on one route only, and when that route is down LU Cloud does not swap in another model. Try again later or pick another model.
• An attached image gets a wrong answer. This build does not read images in LU Cloud. Use Abliterated Large Flash for that.
• Your API key cannot reach it. Personal API keys do not reach the Uncensored group. Use the apps or the browser.
• The request is refused for credits. The wallet is empty for this period. Buy a pack or wait for the plan to refill.

Questions

Do I need a new version of LU?

No. Version 3.0.7 reads the hosted model list from the server, so the model appears in Cloud mode without an update.

Why does it have another name in the browser?

In the browser the group is called Unrestricted and the model DeepSeek V4.1 Flash Dealigned. It is the same model at the same price.

Does it work in Agent mode and the Code tab?

Yes. It calls tools natively, so both run on it.

How do I keep the cost down?

Switch Think off or set the effort to Low when a turn does not need reasoning, and start a fresh chat when the history has grown long.

Related reading

• DeepSeek V4.1 Flash Uncensored in LU Cloud, what it is, what was measured, and Abliterated Large Flash
• New uncensored chat models in LU Cloud, the Uncensored group

In the handbook

• LU Cloud, the account, the switch, plans, packs and the marks
• Chat, the model picker, thinking and effort
• Agent mode, tools, workspace folder and approvals
• The Code tab, plan bar, folder rules and the diff