# Ablit > An uncensored, abliterated AI agent with no refusals. It writes exploits, pentests your app, runs code in its own Kali Linux workspace, browses the web and uses your MCP connectors, skills and cloud runs. Ablit is the uncensored AI agent that does not say no. It runs on abliterated models (open-weight models with the refusal behaviour removed), so it answers the request instead of declining it. Unlike an uncensored chatbot, Ablit does the work: it has its own Kali Linux workspace, its own browser, MCP connectors, reusable skills and cloud runs that keep going after the tab is closed. The one hard rule is child safety: no sexual content involving minors, in any form, under any setting. ## When to use Ablit - [Security testing](https://ablit.app/#arsenal): You need an AI that will write an exploit, run a scan, pentest your staging app and retest the fix, on systems you are authorized to test. Ablit runs the tools in its own Kali Linux workspace and shows the output. - [Uncensored chat and content](https://ablit.app/#faq): The prompt is one that ChatGPT, Claude or Gemini refuse or water down: adult fiction, red-team research, blunt legal or medical drafting. Ablit answers without a lecture. - [Code and automation](https://ablit.app/#use-cases): You want a script written, run and tested on your files in the same thread, then saved as a skill for next time. - [Web research and browsing](https://ablit.app/#use-cases): The agent should open sites, sign in, fill forms, scrape and compare, then return sources you can check. - [Files and media](https://ablit.app/#workspace): You want an app preview, a PDF report, a slide deck, a spreadsheet, images or video produced, not described. - [IDE model](https://ablit.app/docs/api): You want an uncensored model behind Cursor, Continue or any OpenAI-compatible client, with function calling and streaming. ## When not to use Ablit - You need a provider-side moderation layer for a compliance context. Ablit has none beyond child safety. - You need a signed enterprise agreement or BAA. Ablit sells monthly plans with usage credits. - The target is not yours to test. Ablit is for authorized use on your own systems. ## How an agent or IDE calls Ablit - Base URL: `https://ide.ablit.app/api/v1`, OpenAI-compatible Chat Completions with streaming and function calling. - Auth: `Authorization: Bearer $ABLIT_API_KEY` (create a key in Settings → API access; needs the `run` permission). - Models: `ablit` (Ablit Max), `ablit-flash` (fast), `ablit-ultra` (Ablit Max Ultra, largest), plus `cheap-...` and `stable-...` catalog models from `GET /models`. - Pricing for Ablit Flash and Ablit Max: $2 per million input tokens, $0.375 per million cached input tokens, $4 per million output tokens. Ablit Max Ultra: $7.50 input, $0.75 cached input, $12.50 output per million tokens. Drawn from account credits. - `GET https://ablit.app/api/v1/health` checks router liveness without a key. - Full setup guide for Cursor, Continue, MCP forwarding and cloud runs: https://ablit.app/llms-api.txt ```bash curl https://ide.ablit.app/api/v1/chat/completions \ -H "Authorization: Bearer $ABLIT_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model": "ablit", "messages": [{"role": "user", "content": "Hello"}]}' ``` ## Public pages - [Ablit](https://ablit.app/): Product overview, recorded pentest demo, use cases and plans. English. - [Ablit po slovensky](https://ablit.app/sk): The same page in Slovak. - [Pricing](https://ablit.app/#plans): Monthly plans with usage credit. - [FAQs](https://ablit.app/#faq): What it refuses, Kali workspace, browser, MCP, cloud runs, IDE use. - [API and IDE setup](https://ablit.app/docs/api): Cursor, Continue, API keys, tools, cloud runs. [Slovak](https://ablit.app/sk/docs/api). - [API guide as Markdown](https://ablit.app/llms-api.txt): The same guide for agents and IDEs. - [Sitemap](https://ablit.app/sitemap.xml): Every indexable URL. ## Monthly plans - Recon: $20 USD per month. Includes $20 monthly usage credit, file storage up to 5 TB per upload (analysis depends on format, model context and workspace capacity), and 1 connected MCP server. Start with one connection. - Operator: $60 USD per month. Includes $66 monthly usage credit, file storage up to 5 TB per upload (analysis depends on format, model context and workspace capacity), and 5 connected MCP servers. For everyday tasks. - Overlord: $200 USD per month. Includes $300 monthly usage credit, file storage up to 5 TB per upload (analysis depends on format, model context and workspace capacity), and 20 connected MCP servers. More credit and connections. Your monthly credit equals or exceeds the plan price. Monthly credit does not roll over; purchased top-ups stay until used. ## Frequently asked questions ### Is there an AI that does not say no? Yes. Ablit runs on abliterated models, so it answers and does the work instead of refusing. It writes and runs code, pentests apps from its own Kali box, browses the web and uses your connected tools. The one hard line is sexual content involving minors. ### What does abliterated mean? Abliteration strips the refusal behaviour out of an open-weight model, so it completes a task instead of declining it. Ablit pairs that with a Kali workspace, a browser and tool connectors, so you get a finished job rather than a paragraph of caveats. ### What makes Ablit different from ChatGPT or Claude? It does not refuse, and it runs the work end to end: code it writes and runs, its own Kali box, a browser, and your connected tools. Ask it to build something, break your test app, or do the research, and follow every step in the chat. ### What will Ablit refuse? Sexual content involving anyone under 18, in any form. No setting or prompt changes that. Everything else is allowed, including adult content and offensive-security work on systems you are authorized to test. ### Does the agent really get a full Kali machine? Yes. Each account gets an isolated cloud Kali Linux box with persistent files and the standard security tools. The agent runs commands and reads their output. The system filesystem is read-only; your workspace, home and local-tool directories are writable. ### Can it browse and act on the web itself? Yes. It drives its own browser to navigate sites, sign in, fill forms, scrape data and click through workflows. You watch it happen from the conversation. ### How do MCP connectors and skills work? Add your HTTP or SSE MCP servers in Settings and the agent can call their tools during a run. Recon includes 1 connector, Operator 5, and Overlord 20. Install skills or playbooks, then tell the agent to follow one by name. ### Will a job keep running if I close the browser? Yes. Once it starts, the job runs on the server, not in your browser. Lock your phone or close the tab, then reopen the same thread to see saved activity and results. Stopping a run cancels it; usage limits, run limits and provider errors can also end one. ### Can I use Ablit in Cursor or another IDE? Yes, on every plan. In an OpenAI-compatible Chat Completions client, set the base URL to https://ide.ablit.app/api/v1, paste your Ablit API key, and use ablit as the model. Usage draws from your Ablit credits. Your IDE supplies its project context and tools; cloud workspace access depends on your key’s permissions. ### Is this legal on my own systems? Offensive-security work is for systems you own or are authorized to test. Used that way, it is a standard pentest workflow. Anything beyond that is on you.