Mistral

Follow

3/26/2026

Mistral AI has launched Voxtral, a powerful new family of speech-to-text models. Unlike text-to-speech, Voxtral focuses on converting spoken words into text, claiming to match or exceed ElevenLabs' Scribe accuracy at under half the price. Crucially, Mistral has open-sourced the model weights for free commercial use, enabling SMBs to deploy advanced voice processing securely on their own infrastructure.

Mistral

Follow

3/27/2026

Mistral AI has launched Voxtral TTS, its first text-to-speech model. Supporting nine languages and available with open weights, Voxtral TTS allows businesses to run and customize it on their own infrastructure, offering greater control over data, costs, and integration. It's designed for enterprise use, including voice assistants and customer support, even running efficiently on consumer hardware.

Mistral

Follow

3/26/2026

French AI company Mistral has launched 'Voxtral TTS,' a new open-source text-to-speech model. This model runs efficiently on smart devices and offers a low-cost solution, potentially accelerating voice AI adoption for small and medium-sized businesses. It supports nine languages and aims to provide human-like audio for various applications, from customer support to sales engagement.

Mistral

Follow

3/28/2026

Mistral AI has launched Voxtral TTS, its inaugural voice generation model. This open-weight, 4-billion parameter model is designed for low-latency, multilingual speech, supporting nine languages natively. It can also clone voices from just a 3-second audio sample, offering a compelling alternative to existing paid voice APIs and potentially transforming how SMBs use audio.

Is China picking back up the open source AI baton? Z.ai, also known as Zhupai AI, a Chinese AI startup best known for its powerful, open source GLM family of models, has unveiled GLM-5.1 today under a permissive MIT License, allowing for enterprises to download, customize and use it for commercial purposes. They can do so on Hugging Face. This follows its release of GLM-5 Turbo, a faster version, under only proprietary license last month. The new GLM-5.1 is designed to work autonomously for up t...

Mistral

Follow

3/30/2026

French AI firm Mistral has launched Voxtral TTS, an open-source voice generation AI that promises performance on par with industry leader ElevenLabs. This new model is designed to be more efficient and cost-effective, potentially opening up high-quality audio content creation to SMBs who previously faced high technical and financial barriers. It supports multiple languages and offers low-latency processing, making it suitable for a wide range of business applications.

Anthropic

Follow

4/5/2026

A new research, linked to Anthropic's AuditBench, introduces "Probe-Mediated Adaptive Auditing." This method can uncover hidden behaviors in AI models, specifically 56 variants of Llama 3.3 70B, without needing a reference model. It marks a significant step forward in AI safety auditing, offering crucial insights for SMBs leveraging AI.

Anthropic

Follow

4/5/2026

The AI landscape is buzzing with activity, as evidenced by the March 30, 2026, AI Daily Brief. Anthropic's Claude model is rapidly gaining market share, while the company prepares for a highly anticipated IPO. Concurrently, China's DeepSeek AI experienced a service outage, underscoring the dynamic and sometimes volatile nature of the artificial intelligence industry.

MiniMax just open-sourced M2.7, an advanced AI model with self-evolution capabilities. This Mixture-of-Experts (MoE) model is designed to boost productivity in software engineering and office tasks, offering faster processing and lower operating costs. It performs well on complex benchmarks, suggesting potential for significant automation in SMB operations.

OpenAI

Follow

4/3/2026

Arcee has launched Trinity-Large-Thinking, a new open-source AI model developed in the U.S. This model stands out because it's freely downloadable and customizable, giving businesses, especially SMBs, a powerful, domestically sourced option for building tailored AI solutions. It marks a potential shift back towards U.S. leadership in open-source AI.

Mistral

Follow

4/2/2026

Google has released its latest lightweight language model, Gemma 4, under the Apache 2.0 license. This is a big deal because previous Gemma versions came with Google's own restrictive license, making legal review a headache for many businesses. Now, Gemma 4 is truly open source, clearing the way for SMBs and innovators to confidently integrate Google's powerful AI into their products and operations without legal friction.

Anthropic

Follow

5d

Claude Code Code Review, Deepseek V4, Gemma 4, OpenClaw Update, Copilot Cowork, & More! HUGE AI News Fathom Journal