Home / features

The full cockpit, panel by panel.

Everything below runs on your machine, against your local Ollama server. No feature phones home, because there's no home to phone.

Two ways to talk

Generate mode vs. Chat mode

The API dropdown (top center) controls how the app talks to the model. Both are useful; they solve different problems.

Generate — the one-shot workbench

Each request stands alone; the model "forgets" the previous question immediately. This is the mode for tasks, not conversations — and it's the only mode that accepts images and audio. You can optionally carry the context tokens into the next request, which gives you conversation-like continuity even here.

visionaudioscreenshotsone-shot

Chat — the conversation with a superpower

The whole history is kept and sent with every message. The superpower: you can switch the LLM mid-conversation. Brainstorm with a fast 8B model, then hand the thread to a heavyweight for the final answer — the new model inherits the full context.

The bridge: Generate > Chat

Chat mode can't receive images directly, so the app gives you a workflow instead of a wall:

  • In Generate, upload an image and ask about it ("Describe this diagram").
  • When the answer arrives, click Generate > Chat at the top.
  • The question and the analysis are copied into the chat history — now ask follow-ups in Chat mode. The model "knows" the image's content from the transcript.
Where you type

The intelligent prompt area

The bottom of the app is built for real work, not one-liners:

  • It grows with you. Paste a long text and the input expands automatically, then becomes scrollable at max height — big code blocks stay manageable without eating your screen.
  • Attached files are visible. Everything you attach appears as a chip above the input, so you always know exactly what context is being sent to the model.
  • Keys: Enter or the ▶ button sends everything, attachments included. Shift+Enter inserts a line break for structuring your prompt.
Eyes

Vision & screenshots

Upload .jpg or .png files in Generate mode and analyze them with a vision model like llava. Diagrams, error dialogs, whiteboard photos, UI mockups — described, transcribed, or interrogated, without the image ever leaving your disk.

The built-in screenshot button shortens the loop: capture your screen and pipe it straight into the vision model. "What's wrong with this error message?" becomes a two-click question.

Remember: vision lives in Generate mode. Use the Generate > Chat bridge to keep discussing an image conversationally.
Ears

Audio in, locally

Upload .wav or .mp3 files — or hit the record button and speak — and analyze the audio with an audio-capable model like gemma4. Voice memos, meeting recordings, dictation: processed on your hardware, useful even when the content is exactly the kind of thing you'd never upload to a cloud transcription service.

Memory

File reading & the RAG tool

Read files into context

Import .pdf, .txt, and .json files directly into the prompt context. Good for anything that fits in the model's context window: a config file, an article, a spec.

RAG — chat with documents too big to paste

For large PDFs and text files, the RAG Tool builds local knowledge extraction over the document, then answers your questions from the relevant parts. A 200-page contract, a thesis, a technical manual — interrogated on your workstation, indexed on your disk, exposed to no one.

Readable answers

Rendered output: Markdown, HTML, LaTeX, thinking

  • Markdown & HTML views render responses properly — headings, tables, links — with syntax-highlighted code blocks labeled by language.
  • Formula output: LaTeX support renders real mathematics. Ask a stats question, get actual notation instead of ASCII soup.
  • Visible thinking: reasoning models' think-blocks are displayed as distinct callouts, so you can watch the model work — or collapse the noise and read the answer.
  • Export: turn a rendered response into a PDF when you need to keep or share it.
Structured

JSON mode — answers a program can trust

Flip on JSON mode and the model is forced to respond in structured JSON — not "usually JSON, occasionally an apology," but parseable output. This is the feature for people who pipe AI into other things: extraction pipelines, automations, form-fillers, anything downstream of the answer.

Hands

Tools, functions & code execution

Tools & functions

Let the model call local Python through tool/function calls. The model decides it needs a computation or a lookup, invokes the tool, and works the result into its answer — the agent loop, running entirely on your box.

Direct code execution

When a response contains a Python, PowerShell, or Batch script, you can execute it straight from the app: pick the code block, review it, hit Run. The write-copy-paste-debug shuffle collapses into a click.

Sharp tool, by design. Executed code is real code on your real machine. The app puts you in the loop — you see the script before you run it. Read it. That's the deal with owning your tools.
The exclusive

SSH Bot — an autonomous agent in your rack

This is the feature no cloud chatbot will ever be allowed to ship. The SSH Bot turns Cryptic Thinking into an autonomous system administrator: it connects to a server over SSH with your credentials, executes commands, reads the terminal output, and reacts to what it sees on its own — chaining steps until the job is done.

  • "Check disk usage on this box and find what's eating space" — it runs df, sees the numbers, drills into the offending directory, reports back.
  • Diagnose a failing service: it reads the logs, forms a hypothesis, checks it.
  • You watch the live terminal the whole time. You chose the server, you provided the credentials, you can stop it at any moment.
Why local matters here most of all: an AI with SSH access to your infrastructure is exactly the thing you don't route through a third party's cloud. The agent, the credentials, and the terminal output all stay on your machines.
Imagination

Image generation

Generate images locally with x/z-image-turbo and x/flux2-klein. Concept art, placeholders, illustrations — produced by your GPU, with no content policy server deciding what you're allowed to draw and no gallery of your prompts on someone else's servers.

Under the hood

Full parameter control

Cloud UIs hide the dials. Cryptic Thinking hands them to you: temperature, context length, system prompts, and the rest of the model parameters — adjustable per session, with presets for the setups you keep coming back to. The model info view shows what the selected model supports (the capability chips turn green for vision, tools, and friends), so you always know what you're working with.