Talk. Don’t Type.
The first truly voice-native AI assistant for macOS. Natural conversations with powerful local LLMs — completely offline, endlessly private.
While others bolt on voice as an afterthought, Thursday was designed from the ground up for natural voice conversations. Just talk — like you would to a colleague.
Choose from 5 different Whisper models for the perfect balance of speed and accuracy. From tiny (fastest) to large (most accurate).
Natural, human-like responses with Kokoro TTS or system voices. Conversations that feel alive, not robotic.
Create personalities with unique prompts for on-point voice conversations. Be creative — or just practical.
All your conversations saved locally, forever. Go back to any conversation anytime — your data, your control.
Run everything locally with Ollama or LM Studio. Your data never touches any cloud. Ever.
Everything you need in a voice-native AI assistant, nothing you don’t
Integrate with Gemini, ChatGPT, Claude, OpenRouter, DeepSeek, and any OpenAI-compatible endpoint.
Model Context Protocol integration. Access files, browse the web, and extend with custom tools.
Beautiful interface that matches your system preference. Easy on the eyes, day or night.
Connect any OpenAI-compatible API. Use your favorite provider or self-hosted models.
Built with SwiftUI for Apple Silicon. Lightning fast, energy efficient, feels right at home on Mac.
Maintains conversation context for coherent, multi-turn dialogues. Just like talking to a real assistant.
Works with all major AI providers and any OpenAI-compatible API
No subscriptions, no hidden fees. Pay once and use Thursday forever.
Try everything Thursday has to offer with no commitment.
Get lifetime access and support all future development.
Not satisfied? Get a full refund within 30 days, no questions asked.
Everything you need to know about Thursday
Thursday requires a Mac with Apple Silicon (M1, M2, M3, or newer). We recommend at least 16GB RAM for fluid voice interactions with local models.
Yes! When using Ollama or LM Studio, everything runs locally. No internet required, no data leaves your Mac. Your private conversations stay private.
Qwen 3 and Gemma 3 are excellent for fast, natural conversations. Larger models like Llama 3.2 70B give richer responses if you have the RAM.
Absolutely! Connect to Gemini, ChatGPT, Claude, OpenRouter, DeepSeek, or any OpenAI-compatible endpoint alongside local models.
You can activate your license on up to 3 Macs. Need to switch devices? Simply deactivate from the old one first.
After 7 days, you’ll need to purchase a license to continue. All your conversations and settings remain intact.
Stop typing. Start talking. Experience AI the way it should be — natural, private, and powerful.
Get Thursday for $10