raise money, no cap?

Author

Dan Sutera
March 14, 2025 • Estimated Reading Time: 5 minutes

In partnership with

Sesame Labs Releases Open Speech Model

Sesame Labs releases CSM, an open-source speech model trained on one million hours of data. The Apache-licensed system offers emotional intelligence and voice cloning capabilities.

 

Gemma 3: Revolution For AI On Single Devices

The TLDR
Google DeepMind launches Gemma 3, an open AI model that delivers impressive performance while running on single GPUs. The system outperforms larger models, supports 140+ languages, and handles multiple media types with context windows of 128K tokens.

Google DeepMind has introduced Gemma 3, a new generation of open AI models that combine impressive performance with unprecedented portability. As a slimmed-down version of Gemini 2.0 technology, Gemma 3 is the most powerful model that can run on a single GPU or TPU.

What makes Gemma 3 so special is the combination of peak performance and accessibility. In user preference tests, it outperforms larger models such as Llama-405B and DeepSeek-V3 while running on just a single accelerator.

With support for over 140 languages, an impressive 128K token context window, and the ability to process text, images, and short videos, Gemma 3 opens up a whole new range of possible applications. The series is available in different sizes (1B, 4B, 12B, and 27B), allowing developers to choose the appropriate model based on their hardware.

In addition, Google places a high value on responsible development with extensive security protocols and provides an image security check with ShieldGemma 2.

After the great success of the first version of Gemma with over 100 million downloads, Gemma 3 is another important step in Google's effort to make high-quality AI technology accessible and running on the edge!

#AD

Find out why 1M+ professionals read Superhuman AI daily.

In 2 years you will be working for AI

Or an AI will be working for you

Here's how you can future-proof yourself:

  1. Join the Superhuman AI newsletter – read by 1M+ people at top companies

  2. Master AI tools, tutorials, and news in just 3 minutes a day

  3. Become 10X more productive using AI

Join 1,000,000+ pros at companies like Google, Meta, and Amazon that are using AI to get ahead.

 

AI Investor Wires $100K After Brief Call

An AI called No Cap autonomously invested $100,000 in a startup after just a three-minute call. The female-presenting system signs agreements and provides business connections independently.

 

Shuka v1: India's Audio AI Breakthrough

Developers use Meta's Llama to create Shuka v1, India's first open-source audio language model. The breakthrough represents India's entry into audio AI development.