RadostLLM
RadostLLM Pro

Local AI for daily work.

Pro saves every chat, gives you every Ollama option, manages your models and works on your code in agent mode. Like Lite, it runs on your computer with no account and no tracking.

$19.99once, no subscription

US price. The stores set the price for each country and currency, so yours may differ.

  • Mac App StoreLink coming soon
  • Microsoft StoreLink coming soon
Release notes draftOllama
1,284 / 8,192 tokens
Turn these commit messages into short release notes for users.commits.txt
qwen3:8b

Version 2.4

  • Search now finds words inside attachments.
  • Exports keep code blocks and tables.
  • Fixed a crash when a model was removed while loading.
41.8 tokens/s · 64 tokens
Ask anything
Fig. 1. RadostLLM Pro with saved chats, an attachment, a context meter and tokens per second. An illustration with a sample chat.
Price
$19.99 once
Subscription
None
Servers
Ollama, LM Studio, llama.cpp, vLLM
Account
None
Telemetry
None
Platforms
macOS and Windows

What Pro adds

Everything in Lite, plus the tools you reach for when local AI is part of your day.

Keep every chat

Chats stay on your computer, encrypted, and easy to find again.

  • Unlimited saved chats with search, rename, pin and delete
  • Folders and tags, with a tag filter in the sidebar
  • A model and a system prompt for each chat
  • Export a chat to Markdown, PDF or JSON
  • Back up and restore all chats, including Lite backups

Tune the model

Every option Ollama has, without a terminal.

  • temperature, top_k, top_p, num_ctx, seed, repeat_penalty, min_p, num_predict, stop, mirostat and keep_alive
  • Presets saved for each model
  • Thinking on or off for reasoning models
  • A JSON schema per chat for structured output
  • Tokens per second and a context meter for every answer

Manage your models

Download, inspect and unload models from the app.

  • Download models with progress, delete and copy them
  • Model details, and loaded models with their memory use
  • Ollama on this computer or on another machine in your network
  • OpenAI-compatible servers: LM Studio, llama.cpp, vLLM

Work faster

Less retyping, more answers.

  • A prompt library with {{variables}}
  • Knowledge: answers from the text, code and PDF files of a folder, with sources under each answer
  • Text, code and PDF files as attachments, images for vision models
  • Regenerate, edit and resend, or branch a chat from any message
  • Ask two models the same question side by side, then continue either answer
  • Quick Ask: a small window on Cmd or Ctrl + Shift + Space, plus a menu bar or tray icon
  • Lock the app with Touch ID on macOS

Code with it

A local model that can read your project, with you in charge of every change.

  • Agent mode: the model lists, reads and searches the files of a folder you pick
  • git status, diff and log, and a Review git diff action for your uncommitted changes
  • File writes show a diff first, unless you let a chat apply them; shell commands and web downloads wait for your click
  • Local MCP servers in the usual mcpServers format, each tool call approved
  • Save any code block to a file

Coming from Lite

Pro is a separate app, so Lite keeps working. Bring your chats along with a backup.

  1. Step 1In RadostLLM Lite, open Settings and choose Back up. Save the file.
  2. Step 2In RadostLLM Pro, open Settings and choose Restore.
  3. Step 3Pick the backup file. Your Lite chats appear in the sidebar.