Version 1.9.0: Going full speed!
We made Alter a lot faster and integrated the world's fastest inference provider. Say hello to Cerebras!
Hey there,
This version improves the speed of text rendering. Plus, we have integrated Cerebras, the world's fastest inference provider on the planet, generating 2000 tokens per second on Llama 3.1!
With these new models and changes, Alter feels instant:
And we’re proud to say that Alter is faster than ChatGPT desktop at rendering replies:
We have 3 others improvements:
- You can copy previous assistant replies
- You can choose system, dark, or light mode from the settings
- You can choose the font size of the Chat in the new Appearance tab in Preferences
Cheers,
Sam - Founder
New video:
Other videos:
- 5 features of Alter in 2 minutes ⚡
- Summarize Youtube Videos: What's faster, Alter or ChatGPT?
- How to generate images with Ideogram without prompting
- Chat with your emails on Apple Mail
- Write in your style with Alter
- 📺 Repurpose content with Alter: A step-by-step guide
- Chat with a Reddit Page Using Alter MacOS App
New features:
- New provider: Cerebras
- New models: Llama 3.1 70B, Llama 3.1 8B, Nemotron 70B and more
- Customization: Dark and Light mode from the settings, custom Chat font-size
- New browser: support for Microsoft Edge
- Chat: Copy past assistant replies
Improvements & Bug fixes:
- Streaming and parser: performance improvements
- Offline mode: embed font awesome, Katex, Texmath
- Reminder: fix dark mode support
