Vresk
Product Hunt Kennedy Booth
Vresk is a new AI workspace that picks the model for each task. Your notes and memory stay put, so you can switch models without starting over.
Based on reporting by Product Hunt, Kennedy Booth — read the original for the full story.
Summary, retelling and take written by AI under human oversight; images are AI-generated illustrations. How we work · Report an error
Vresk is launching as an AI workspace built around open models, with names like Kimi, DeepSeek, Qwen and Llama in the mix. The pitch is simple: you ask, and the workspace chooses the model it thinks fits best for that job.
That matters because model-hopping is usually messy. Here, the notes, history and memory live inside Vresk, not inside one model’s little box. So if the app sends one prompt to a different model next time, you’re not dragged back to square one.
The product also tries to cover more than plain chat. Users can ask for an answer, a live widget or a designed deck, which suggests Vresk wants to be a working surface, not just a prompt box with nicer branding.
And it makes the model choice visible. Every reply says which model wrote it, which is a small but useful bit of honesty in a category that often hides the plumbing until something breaks.
Vresk is free to try and then costs $20 a month flat. That pricing, plus the focus on open models, puts it squarely in the camp of tools that want flexibility without making people feel like they need a subscription spreadsheet just to think.
My take — AI-written commentary, not fact-checked reporting
The cleanest idea here is also the least flashy: let the workspace do the switching, not the user. That’s the right answer to model sprawl, and it’s a nice little rebuke to products that treat vendor lock-in like a feature. Showing which model answered is basic dignity, really.
Read more about this at: Product Hunt