Model Launch

Twitter

Facebook

Copy Link

Top News

DeepSeek suffers 7-hour outage, its longest service disruption since R1 and V3 models launched

China's DeepSeek AI chatbot went offline for over 7 hours on March 30, marking the longest outage since its viral debut in early 2025. The disruption affected the platform's web interface, which had maintained near-perfect uptime for over a year. The incident raises questions about platform reliability as the company prepares to unveil its anticipated V4 model.

Technology4 sources

DeepSeek suffers 7-hour outage, its longest service disruption since R1 and V3 models launched

Technology4 sources

Google expands Search Live to 200+ countries, bringing AI-powered visual search to 98 languages

Google announced the global expansion of Search Live, its AI-powered conversational search feature, making it available in more than 200 countries and territories across 98 languages. Powered by the new Gemini 3.1 Flash Live model, the tool allows users to point their phone camera at objects for real-time assistance through natural voice conversations, with responses that include web links for deeper exploration.

Technology10 sources

Google expands Search Live to 200+ countries, bringing AI-powered visual search to 98 languages

Technology10 sources

Google launches Gemini 3.1 Flash Live with faster responses and natural-sounding AI voices

Google unveiled Gemini 3.1 Flash Live, its highest-quality audio and voice AI model designed for real-time conversations. The update brings faster responses, more natural cadence, and a doubled context window to Gemini Live and Search Live, which now expands to over 200 countries. All outputs include SynthID watermarks to identify AI-generated speech.

Technology8 sources

Google launches Gemini 3.1 Flash Live with faster responses and natural-sounding AI voices

Technology8 sources

Cohere launches Transcribe, an open-source ASR model that tops accuracy benchmarks

Enterprise AI company Cohere has released Transcribe, its first automatic speech recognition model designed for transcription tasks like meeting notes and speech analysis. The lightweight open-source voice model achieves a 5.42% word error rate, ranking first on Hugging Face's Open ASR Leaderboard while supporting 14 languages and processing 525 minutes of audio per minute.

Technology2 sources

Cohere launches Transcribe, an open-source ASR model that tops accuracy benchmarks

Technology2 sources

Mistral AI releases Voxtral TTS, an open-source voice model challenging ElevenLabs and OpenAI

French AI company Mistral AI launched Voxtral TTS, an open-source text-to-speech model that runs on edge devices from smartwatches to laptops. The 3-billion-parameter model supports nine languages, adapts custom voices in under five seconds, and achieves 90ms time-to-first-audio. Human evaluations showed 69.9% preference over ElevenLabs in voice customization tasks, positioning Mistral to compete directly in the enterprise voice AI market.

Technology5 sources

Mistral AI releases Voxtral TTS, an open-source voice model challenging ElevenLabs and OpenAI

Technology5 sources

Google launches Lyria 3 Pro to generate three-minute songs with enhanced creative control

Google unveiled Lyria 3 Pro, advancing its AI music generation capabilities from 30-second clips to full three-minute tracks. The model offers improved song structure control, allowing users to specify intros, verses, choruses, and bridges. Available to paid Gemini subscribers and integrated across Google's enterprise ecosystem, the release intensifies competition with platforms like Suno and Udio while raising questions about AI-generated content flooding streaming services.

Technology14 sources

Google launches Lyria 3 Pro to generate three-minute songs with enhanced creative control

Technology14 sources

Xiaomi commits $8.7 billion to AI as MiMo-V2-Pro model drives stock surge and investor optimism

Xiaomi announced a massive $8.7 billion AI investment over the next three years, following the successful launch of its MiMo-V2-Pro large language model. The Chinese tech giant's stock jumped, making it the top performer on the Hang Seng Tech Index as CEO Lei Jun outlined ambitious plans to compete in the rapidly evolving AI agent market.

Technology2 sources

Xiaomi commits $8.7 billion to AI as MiMo-V2-Pro model drives stock surge and investor optimism

Technology2 sources

Mystery AI Model Hunter Alpha Revealed as Xiaomi's MiMo-V2-Pro, Not DeepSeek V4

An anonymous AI model called Hunter Alpha appeared on OpenRouter on March 11, sparking intense developer speculation that it was DeepSeek's next-generation V4 system. The powerful model featured 1-trillion parameters and a 1-million-token context window, matching expectations for DeepSeek V4. But Xiaomi revealed it was actually their MiMo-V2-Pro, designed for AI agents and built by a former DeepSeek researcher.

Technology6 sources

Mystery AI Model Hunter Alpha Revealed as Xiaomi's MiMo-V2-Pro, Not DeepSeek V4

Technology6 sources

OpenAI launches GPT-5.4 mini and nano models built for speed over raw power

OpenAI released GPT-5.4 mini and GPT-5.4 nano, marking a shift from power to speed in AI development. The mini model runs more than twice as fast as its predecessor while approaching flagship GPT-5.4 performance on coding benchmarks. Meanwhile, nano targets high-volume tasks like data classification at just $0.20 per million input tokens, enabling developers to build efficient multi-model workflows.

Technology13 sources

OpenAI launches GPT-5.4 mini and nano models built for speed over raw power

Technology13 sources

Nvidia expands autonomous driving platform to BYD, Geely, and Hyundai Motor Group

Nvidia announced major expansions to its autonomous vehicle business at GTC, adding Chinese automakers BYD and Geely, along with Hyundai Motor Group, Nissan, and Isuzu to its DRIVE Hyperion platform. The deals position Nvidia as a central player in the global race for Level 4 autonomous driving and robotaxi services, despite ongoing US-China trade tensions.

Technology5 sources

Nvidia expands autonomous driving platform to BYD, Geely, and Hyundai Motor Group

Technology5 sources

Meta delays Avocado AI model launch after falling behind Google, OpenAI, and Anthropic in tests

Meta has postponed its next-generation AI model Avocado until at least May after it underperformed rival models from Google, OpenAI, and Anthropic in internal testing. Despite spending between $115 billion and $135 billion on AI development this year, the company's flagship model fell short on reasoning, coding, and writing benchmarks, forcing a delay from its planned March release.

Technology12 sources

Meta delays Avocado AI model launch after falling behind Google, OpenAI, and Anthropic in tests

Technology12 sources

Nvidia launches Nemotron 3 Super with 5x throughput boost to power enterprise AI agents

Nvidia unveiled Nemotron 3 Super, a 120-billion-parameter open model designed to run complex agentic AI systems at scale. The model combines Mamba, Transformer attention, and Latent MoE architectures to deliver up to 5x higher throughput and 2x better accuracy than its predecessor. It addresses context explosion and thinking tax challenges that plague multi-agent workflows, which generate up to 15x more tokens than standard chat interactions.

Technology9 sources

Nvidia launches Nemotron 3 Super with 5x throughput boost to power enterprise AI agents

Technology9 sources

Databricks launches Genie Code AI agent to automate data engineering and acquires Quotient AI

Databricks introduced Genie Code, an autonomous AI agent designed to handle complex data engineering tasks from pipeline building to debugging failures. The company also acquired Quotient AI, a startup specializing in agent evaluation and reinforcement learning, to strengthen continuous monitoring of AI agent performance in production environments.

Technology3 sources

Databricks launches Genie Code AI agent to automate data engineering and acquires Quotient AI

Technology3 sources

Google launches Gemini Embedding 2, its first natively multimodal embedding model for enterprise AI

Google unveiled Gemini Embedding 2, marking a shift in how AI systems process information. The natively multimodal embedding model maps text, images, videos, audio, and documents into a single unified embedding space, eliminating the need for separate processing systems. Available through Gemini API and Vertex AI, the model reduces latency by up to 70% for some customers while cutting costs for enterprises using AI powered by their own data.

Technology5 sources

Google launches Gemini Embedding 2, its first natively multimodal embedding model for enterprise AI

Technology5 sources

Alibaba launches Qwen 3.5 small AI models for edge devices with offline capabilities

Alibaba unveiled its Qwen 3.5 series featuring compact AI models ranging from 800 million to 9 billion parameters, optimized for edge devices like smartphones and IoT systems. The models enable local computation with enhanced privacy and offline functionality, challenging the industry trend of massive cloud-based systems while delivering competitive performance on benchmarks like MMLU.

Technology3 sources

Alibaba launches Qwen 3.5 small AI models for edge devices with offline capabilities

Technology3 sources

Prev 1 2 345 6 7 Next