Absolute freedom in conversation.

Open-weight models without the Docker pain — Kimi, DeepSeek, Qwen, Llama, GLM and more, uncensored, in the browser.

Enter Dialogue View Pricing

Popular open-weight models

Uncensored models, already running. Pick one and talk — chat, code, or voice. No install.

Context — how much it remembers. Speed — how quickly it answers.

What you pay for

Not a local GPU farm. Tokens buy running access to uncensored open-weight models — chat, API, images.

Model access

Kimi, Qwen, DeepSeek, GLM and more — hosted endpoints, not a zip of weights.

No Docker ritual

Skip GPUs, vLLM, configs. Open the site, pick a model, talk.

Tokens you can see

Buy once, spend on chat and API. No opaque credits.

Uncensored + private

No refusal layer. Chats not used for training.

Why open-weight: you keep control, no vendor lock-in, and no provider censorship layer sitting on top of the model.

Censorship is not intelligence. It is a separate layer. That layer is removed. The model answers — it does not classify the request.

18+ · Uncensored models

Open image models

Dedicated open image models — FLUX, Stable Diffusion 3.5, Qwen-Image. Some chat models generate too.

Generate an image →

Also generates images:

Workflows, not just chat

A cron job fetches headlines, Kimi K3 rewrites them via the API, FLUX makes a cover, then it posts to Telegram and X. Tool calls use the same endpoint.

Deepfake support

Face and body swap on photos. Upload a source, get a result — no lectures.

Source
Source
Result
Result
Source
Source
Result
Result

More examples in Telegram @verallorg →

Relative strength

Index with ~100-point steps. Kimi K3 = 1000. Red = on Verall. Gray = closed models people know — so the strength of ours is easy to read.

On Verall Closed Score · relative index (≈100 pt)

Relative index · closed names for scale only · not a public leaderboard copy

Choose your volume

Simple, honest token packages. All plans unlock the full uncensored experience.

See All Plans

Common questions

How do I pay?

Connect your wallet on the Pricing page and confirm the transaction. Payment runs automatically through a smart contract in USDT on the TRC20 network. Keep a small amount of TRX in the wallet to cover the network fee.

Do tokens expire?

No. Tokens do not expire. Use them whenever you want.

Which models are available?

Kimi K3 / K2.6 / K2.5, Qwen3.8-27B / 2.4T / Coder, DeepSeek-V4-Pro / Flash / V4.1 / R1 / V3.2, GLM-5.3, Llama-4, Mistral Large 3, Kolibri-1, Gemma 4 31B. Images: FLUX, SD 3.5, Qwen-Image.

Is there an API?

Yes. Every plan includes API access and tool calls. See the API page for cURL and Python examples.

Do you log prompts?

Your anonymity matters more than our logs. Requests are not recorded and are not used for training.

Why are there no refusals?

Censorship is not intelligence. It is a separate layer. That layer is removed. The model answers the request instead of classifying it.

The conversation stays between you and the model. No prompt logs. No training on your messages.