Manus launches a cloud-based general AI agent
Manus asks users for an outcome, then plans, researches, codes, operates websites, and returns finished artifacts from a cloud virtual machine.
SEARCH
Search English reports
Browse published English reports with original sources and content provenance.
AI-generated
Alibaba Cloud releases QwQ-32B
Alibaba Cloud releases an open medium-scale model that uses reinforcement learning to improve logical computation and reasoning.
OpenAI releases GPT-4.5
OpenAI releases an improved GPT-4 model with stronger knowledge, generalization and safety performance.
Inception Labs releases Mercury
Inception Labs releases a foundation model built around an alternative architecture and network design.
Alibaba releases Wan2.1
Alibaba's Wan team releases inference code and model weights for Wan2.1, covering text-to-video, image-to-video and other video tasks.
Anthropic releases Claude 3.7 Sonnet
Anthropic releases an upgrade that can switch between standard responses and extended reasoning mode.
xAI begins rolling out Grok-3 beta
xAI began rolling out Grok-3 beta with Think and DeepSearch; the initial release was limited to Grok products, with API access still pending.
Deepgram releases Nova-3
Deepgram releases Nova-3 for real-time and batch transcription, with stronger handling of noise, overlapping speech and word-level timestamps.
Perplexity AI releases Perplexity Sonar
Perplexity fine-tunes its internal core model for its own AI search and precise question-answering workflows.
OpenAI releases o3-mini
OpenAI publicly released o3-mini on January 31, 2025, through ChatGPT and the API.
Hugging Face launches the Open-R1 reproduction project
Hugging Face launches an open community project to reproduce R1 and fill in DeepSeek's unpublished training data and complete training pipeline.
DeepSeek tops the U.S. App Store and jolts AI markets
DeepSeek passes ChatGPT on the U.S. App Store's free chart as Nvidia loses roughly $593 billion in market value during an AI-sector selloff.
Alibaba Cloud releases Qwen2.5-VL
Alibaba Cloud’s Qwen team released the Qwen2.5-VL vision-language model family in 3B, 7B, and 72B sizes, publishing both base and instruction-tuned models.
OpenAI previews Operator, an agent that controls a browser
Operator uses a visual computer-use model to click, type, and scroll through web tasks, handing control back for logins, payments, and sensitive actions.
DeepSeek releases the open-weight reasoning model R1
DeepSeek publishes R1, R1-Zero, six distilled models, and its reinforcement-learning recipe, pushing open reasoning models into the frontier conversation.
ByteDance Seed launches the Doubao Realtime Voice Model
ByteDance Seed launches the Doubao Realtime Voice Model and makes it available in the Doubao app.
DeepSeek launches a free consumer app before R1
DeepSeek puts V3, web search, a reasoning mode, and file analysis into a free mobile app five days before the R1 release.
Mistral AI releases Mistral Small 3.0
Mistral AI releases a third-generation lightweight foundation model optimized for performance on limited hardware.
DeepSeek-V3 brings an efficiency-first open model to the frontier race
DeepSeek releases a 671B MoE model with 37B active parameters, open weights, low API prices, and unusually detailed efficiency disclosures.
Google DeepMind releases Gemini 2.0 Flash Thinking
Google DeepMind releases a derivative of the 2.0 Flash architecture with explicit reasoning for logic-heavy tasks.