Cisco Releases Safety Updates for Actively Exploited SD-WAN Supervisor Flaw
Ravie LakshmananJun 16, 2026Vulnerability / Community Safety Cisco has launched safety updates for a medium-severity safety flaw in Catalyst SD-WAN ...
Tag: Releases
Google Releases Gemini-SQL2: Gemini 3.1 Professional Textual content-to-SQL Scores 80.04% on BIRD Single-Mannequin Leaderboard
Google Analysis staff has introduced the launch of Gemini-SQL2 on X. They described this technique as a breakthrough text-to-SQL functionality ...
Microsoft Analysis Releases Webwright: A Terminal-Native Internet Agent Framework That Scores 60.1% on Odysseys, Up from Base GPT-5.4’s 33.5%
Most internet brokers at present drive a browser one motion at a time. The mannequin receives the present web page ...
LightSeek Basis Releases TokenSpeed, an Open-Supply LLM Inference Engine Concentrating on TensorRT-LLM-Stage Efficiency for Agentic Workloads
Inference effectivity has quietly develop into one of the crucial consequential bottlenecks in AI deployment. As agentic coding methods corresponding ...
Google AI Releases Auto-Diagnose: An Massive Language Mannequin LLM-Based mostly System to Diagnose Integration Check Failures at Scale
When you have ever stared at hundreds of traces of integration check logs questioning which of the sixteen log recordsdata ...
TinyFish AI Releases Full Net Infrastructure Platform for AI Brokers: Search, Fetch, Browser, and Agent Beneath One API Key
AI brokers battle with duties that require interacting with the stay net — fetching a competitor’s pricing web page, extracting ...
TII Releases Falcon Notion: A 0.6B-Parameter Early-Fusion Transformer for Open-Vocabulary Grounding and Segmentation from Pure Language Prompts
Within the present panorama of laptop imaginative and prescient, the usual working process includes a modular ‘Lego-brick’ strategy: a pre-trained ...
Salesforce AI Analysis Releases VoiceAgentRAG: A Twin-Agent Reminiscence Router that Cuts Voice RAG Retrieval Latency by 316x
On this planet of voice AI, the distinction between a useful assistant and an ungainly interplay is measured in milliseconds. ...
Mistral AI Releases Voxtral TTS: A 4B Open-Weight Streaming Speech Mannequin for Low-Latency Multilingual Voice Era
Mistral AI has launched Voxtral TTS, an open-weight text-to-speech mannequin that marks the corporate’s first main transfer into audio era. ...
Google Releases Gemini 3.1 Flash Dwell: A Actual-Time Multimodal Voice Mannequin for Low-Latency Audio, Video, and Software Use for AI Brokers
Google has launched Gemini 3.1 Flash Dwell in preview for builders by means of the Gemini Dwell API in Google ...
Trending.
