Technology News
545 articles
AI technology breakthroughs, research papers, patents, and technical innovations from leading AI startups.
OpenAI is reportedly developing an AI smartphone with Qualcomm and MediaTek for a 2028 launch, with...
OpenAI is reportedly developing an AI smartphone with Qualcomm and MediaTek for a 2028 launch, with Luxshare handling design and production, according to analyst Ming-Chi Kuo. The device will feature...
XiaomiMiMo releases MiMo-V2.5-Pro, an open-source 1.02 trillion parameter MoE language model
China's XiaomiMiMo has released MiMo-V2.5-Pro, an open-source mixture-of-experts model with 1.02 trillion total parameters and a 1 million-token context window. The model is available on Hugging Face....
OpenAI publishes AGI development framework with five principles
OpenAI has released a five-principle framework for AGI development, outlining commitments to responsible scaling, broad distribution of benefits, and avoiding concentration of AI power. The framework...
DeepSeek V4 adopted by AI agent OpenClaw amid Huawei chip collaboration scrutiny
OpenClaw has added DeepSeek's V4 Flash as its default model and also integrated V4 Pro, alongside features like Google Meet support, in an update that optimizes multi-step task consistency. The adopti...
Anthropic overtakes OpenAI in enterprise AI adoption, Ramp Index shows
Anthropic has overtaken OpenAI in enterprise AI adoption, according to Ramp's March 2026 AI Index. The financial data platform reported that Anthropic's enterprise adoption rate surged to 24.4% (up 4....
Anthropic confirms Claude 'dumbing down' bug, resets quotas
Anthropic published a postmortem admitting three bugs caused Claude performance degradation over the past two months: a silent reduction of Claude Code's reasoning level from high to medium on March 4...
Former DeepSeek core member Ruan Chong joins Yuanrong Qixing as chief scientist, details 40B VLA base model
Ruan Chong, a core member of DeepSeek who contributed to VL, V3, R1, and V4, has joined autonomous driving company Yuanrong Qixing as chief scientist. In his debut at the Beijing Auto Show, he detaile...
Meta unveils four new MTIA (Meta Training and Inference Accelerator) chips for AI inference, develop...
Meta unveils four new MTIA (Meta Training and Inference Accelerator) chips for AI inference, developed in partnership with Broadcom. The chips are designed to reduce reliance on external GPU suppliers...
xAI Launches grok-voice-think-fast-1.0, Topping τ-voice Bench at 67.3%
xAI has released grok-voice-think-fast-1.0, a new flagship voice model purpose-built for complex, multi-step voice workflows. According to xAI's benchmark results, the model scores 67.3% on the τ-voic...
DeepSeek unveils V4 model with low-cost high-performance AI strategy
DeepSeek has released a preview version of its V4 large language model, available in Pro and Flash variants, continuing its open-source approach. The company claims the model achieves strong agent-bas...
%20language%20model%20with%201.6%20trillion%20total%20parameters%20and%2049%20billion%20activated%20parameters.%20It%20features%20a%20hybrid%20attention%20architecture%20combining%20Compressed%20Sparse%20Attention%20(CSA)%20and%20Heavily%20Compressed%20Attention%20(HCA)%2C%20achieving%202...&logoUrl=https%3A%2F%2Ffiles.readme.io%2F9294124135914fb6f7626bb3920389713ffaefcb0df8c379cc098cf03ed6796e-small-NVIDIA_Logo_For_LightBG.png&color=%23000000&variant=light){kind=logo}
DeepSeek releases DeepSeek-V4-Pro, a 1.6-trillion-parameter MoE model on NVIDIA NIM
DeepSeek has released DeepSeek-V4-Pro, a 1.6-trillion-parameter mixture-of-experts model, now available via NVIDIA NIM for inference deployment. The model joins DeepSeek-V4-Flash and others in the NVI...
OpenAI Updates Platform with Domain Allowlist Security Setting
OpenAI has introduced a new domain allowlist security setting for its platform, now available in the organization security settings. The feature allows administrators to restrict API access to approve...
Google announces eighth-generation TPUs: TPU 8t and TPU 8i for agentic era
At Google Cloud Next, Google unveiled its eighth-generation Tensor Processing Units, introducing two specialized chips: the TPU 8t designed for large-scale model training and the TPU 8i optimized for...
ThinkingAI Releases Agentic Engine for Enterprise Workflow Automation
South Korean AI company ThinkingAI has published its 'AI Trend 2026' report and launched the Agentic Engine, a platform designed to integrate and manage enterprise work processes. The company emphasiz...
OpenAI releases GPT-5.5, topping all benchmarks and surpassing Opus 4.7
OpenAI unveiled GPT-5.5 and GPT-5.5 Pro on April 24, achieving top scores across all major benchmarks including programming, reasoning, math, and agent tasks. The model outperforms Anthropic's Claude...
DeepSeek releases V4-Pro and V4-Flash models with multi-chip support
DeepSeek today unveiled DeepSeek-V4-Pro, a 1.86-trillion-parameter flagship model, and DeepSeek-V4-Flash, a 284-billion-parameter efficient MoE model. The V4-Flash uses hybrid attention mechanisms (CS...
Huawei backs DeepSeek V4 with Ascend AI chips
Huawei has partnered with DeepSeek to support the training and inference of its next-generation foundation model, DeepSeek V4, using Huawei's Ascend AI chips. The collaboration marks a significant int...
DeepSeek unveils V4 Preview with stronger agent capabilities and 1M-token context, as reports emerge...
DeepSeek unveils V4 Preview with stronger agent capabilities and 1M-token context, as reports emerge of potential funding from Tencent and Alibaba at a $200 billion+ valuation. Why it matters: DeepSee...
OpenAI Retires GPT-4o, GPT-5; Rolls Out GPT-5.3 and GPT-5.5 in ChatGPT
OpenAI has retired GPT-4o, GPT-4.1, GPT-4.1 mini, o4-mini, and GPT-5 from ChatGPT as of February 13, 2026, replacing them with GPT-5.3 as the default model and GPT-5.5 (Thinking and Pro) for advanced...
DeepSeek V4 launches on Huawei Ascend, opens first funding round amid talent exodus
DeepSeek has released V4, a MoE language model with 1.6T total parameters and 1M context window, trained on Huawei Ascend rather than Nvidia GPUs after a major training failure in mid-2025. The model...