After a period of dedicated development, we’re thrilled to announce Haye 2.0! With this significant upgrade, we aim to bring you an even better, more personalized product experience. Let’s dive right into what’s new:
A Fresh, New Interface

Haye 2.0 boasts a completely revamped interface! From the toolbar to the main functional areas, we’ve redesigned everything. To perfectly align with the macOS visual style, we’ve poured our hearts into the interface design, hoping you’ll be captivated by its elegance the moment you open Haye.
Beyond the overall style refresh, we’ve also meticulously refined many of the smaller components, such as the new search box and the enhanced Markdown syntax parsing. We believe that as you use it, you’ll gradually appreciate our dedication to excellence and the care we’ve put into every detail.

To better suit various usage scenarios, Haye 2.0’s main interface supports multiple size modes. You can use it in full-screen immersive mode, or shrink the window to collaborate with other applications. No matter how you switch, we’ve prepared smooth and natural transition animations to make your AI conversations even more seamless.
Embracing Custom Models

In Haye 1.0, we provided a default AI service. But in 2.0, we’re unlocking the power of custom models for you! Simply enter your API key in the settings, and you can experience the powerful capabilities of the world’s leading large language models.
We’ve pre-configured several popular providers for you, including OpenAI, Claude, Gemini, DeepSeek, and Grok. In the future, we’ll also introduce more options like GitHub and OpenRouter. In addition to the latest models we provide, you can freely add your own custom models and start using them in conversations.
Of course, we won’t just haphazardly integrate any model provider and turn it into a mishmash. We’ll carefully evaluate each one to ensure its capabilities and user experience meet our standards before gradually rolling out more choices.
Personalized Configuration That Understands You Better

To enable large language models to understand you better, personalized configuration is crucial. Haye 2.0 introduces a unified rules setting, allowing you to restrict any model and have it respond according to your rules.
We’ve also launched the “Style” feature. You can edit your own style and preferences and enable them at any time during a conversation. With “Style,” you can customize your exclusive prompts more conveniently and naturally.
What’s more, we’re actively developing a memory function that allows the model to gradually learn about you and remember your characteristics through ongoing conversations. (This feature will be added in future iterations.)
A New Pricing Strategy
Haye 2.0 offers two flexible pricing models:
- Software Buyout: If you like Haye’s software features, you can choose to pay once and use it for life.
- Subscription Service: If you also need the model services provided by Haye, you can choose the subscription version. In addition to all the software features, the subscription version also allows you to use the model services provided by Haye. We hope to provide a more convenient and stable AI experience for users who want to get everything in one go, so you don’t have to worry about choosing models.
Free, Powerful Search Mode

In Haye 2.0, our intelligent search mode is completely free! Even if you buy out the software, you can still use our free search service to retrieve content.
Combined with our carefully designed presentation, you can find information more easily. Just click “Search,” enter the keywords, and you’re good to go! (Note: If the current model does not support Function calling, the search model cannot be started, and you need to switch to a model that supports this function.)
A Lighter Toolbar

Haye 1.0’s toolbar was eye-catching, and this design has led the update trend of similar products. In version 2.0, we continue to optimize the performance of the toolbar, making it simpler and more efficient.
When selecting text or images, you can wake up Haye at any time and get convenient services on the toolbar, allowing you to make full use of the powerful capabilities of large language models when editing, creating, and reading.
Sincerity in Every Detail
Haye 2.0 is a huge update, and we’ve invested a lot of time and energy in its development. It’s impossible to list all the improvements here, and we hope you’ll slowly appreciate them as you use it.
More importantly, based on the new framework of 2.0, Haye will have unlimited possibilities. In the coming days, we will continue to iterate on version 2.0 to bring you more AI Chat functions that meet the needs of the times. Stay tuned!