Settings

All settings are stored locally in your browser — nothing is sent to our servers.

AI Behaviour

Controls how the AI thinks and responds in chat.

Default Response Mode

Normal is instant (8B model). Extended Thinking is slower but more structured (49B model).

Stream Responses by Default

Shows text appearing token-by-token as the AI generates it.

Conversation History Length

How many past messages are sent to the AI for context. Higher = smarter but slower.

Task Defaults

Default settings for browser automation tasks.

Default Task Speed

Instant = fast, minimal model calls. Normal = balanced. Thorough = careful, uses larger model for each step.

Default Task Timeout

Maximum time the AI agent is allowed to work on a task before auto-cancelling.

After Task Summary

What the AI shows you after completing a task.

Interface

Visual preferences for the TaskFlow interface.

Auto-Show Browser on Task Start

Automatically opens the live browser popup when you start a task.

Collapse Action Messages

Shows agent step updates in a compact format instead of full bubbles.

Show Message Timestamps

Display the time each message was sent.

Memory & Conversations

Conversations are stored locally in your browser only. Clearing your browser data will delete all conversation history.

Auto-Name Conversations

AI automatically generates a short title based on your first message.

Clear All Conversations

Permanently delete all stored conversations from this browser.

About TaskFlow

Version and legal information.

Version1.0.0
AI ModelsNVIDIA NIM — Nemotron, Llama 3.x
Browser EnginePlaywright / Chromium
BackendFastAPI on Render
FrontendNext.js on Vercel
Terms of Service & Privacy Policy Help & Documentation Source Code (GitHub)