Flash

The news, the context, and what changes.

RSS
Flash

Google Launches Lightweight AI Image Generation Model Nano Banana 2 Lite

Google has released Nano Banana 2 Lite, a new lightweight AI image generation model focused on speed and batch processing. It generates images in just four seconds at a cost of $0.034 per thousand. Available on Google AI Studio, Gemini API, and enterprise platforms, it is designed for high-frequency content creation and will replace the original Nano Banana.

88Z AI editorial1 min read
Flash

Anthropic Launches AI Research Platform Claude Science

Anthropic has released Claude Science, an AI work platform for researchers that offers an integrated environment for literature analysis, multi-step research execution, and result generation, ensuring outputs are auditable and reproducible.

88Z AI editorial1 min read
Flash

AI Chip Startup Etched Completes Tape-Out, Secures $800 Million in Funding

Transformer-specific AI chip startup Etched has completed the tape-out of its A0 chip, securing $800 million in funding and $1 billion in customer orders. Founded by a Harvard dropout, the company is backed by AI luminaries including Karpathy, Li Fei-Fei, and Hinton. Etched has launched its first rack product, utilizing Low Voltage Inference and Cluster-Scale Memory technologies optimized for trillion-parameter MoE model inference, with shipments planned for this summer.

88Z AI editorial1 min read
Flash

Meituan Open-Sources New Trillion-Parameter Large Model LongCat-2.0

Meituan has officially open-sourced LongCat-2.0, a large model with 1.6 trillion total parameters and 48 billion average activated parameters. It is the first model in the industry to complete full-process training and inference on a domestic computing cluster of 50,000 cards, natively supporting 1M long context and optimized for Agentic Coding tasks.

88Z AI editorial1 min read
Flash

Tencent Video Launches Industrial-Grade AI Comic Drama Production Platform WorkRally

Tencent Video has launched WorkRally, an industrial-grade AI platform for producing comic dramas. It targets 2D/3D animation and AI-generated human-like dramas, covering the entire workflow from script analysis to asset management. The platform features an S+ level film and animation skill library, supports expert-level Agent generation and intelligent pipeline scheduling, and includes capabilities for character consistency management and error detection.

88Z AI editorial1 min read
Flash

Tsinghua-affiliated startup Liqing Intelligence completes seed round worth hundreds of millions

Tsinghua-affiliated startup Liqing Intelligence has completed a seed funding round worth hundreds of millions, with participation from Shunwei Capital, HongShan China, and Gaoling Ventures. Founded by Tsinghua AI Institute assistant professor Li Yiming, the team has an average age of 23. The company positions itself as a Physical AI infrastructure provider, developing proprietary data pipelines and differentiable physics engines to achieve a Real-to-Sim-Real loop, requiring only 1% of real-world

88Z AI editorial1 min read
Flash

Huawei Open Sources 92B Parameter Model openPangu-2.0-Flash

Huawei has announced the open-source release of its 92-billion parameter openPangu-2.0-Flash model. Built on Ascend native technology, it aims to provide foundational support for agent applications. Model weights, inference code, and key operators are now available on open-source platforms, with the Pro version expected in July and additional components rolling out in the second half of the year.

88Z AI editorial1 min read
Flash

Global Fintech Platform Airwallex Completes $320 Million Series H Financing

Global fintech platform Airwallex announced the completion of a $320 million Series H financing round, reaching a post-money valuation of $11 billion. Led by Addition, the funds will support AI product development, global compliance, and team expansion, alongside the launch of two new AI products.

88Z AI editorial1 min read
Flash

OpenAI Launches Latest GPT-5.6 Large Language Model Series

OpenAI has released a limited preview of the GPT-5.6 series, featuring Sol, Terra, and Luna models. Sol enhances professional capabilities with new deep reasoning and multi-agent modes; Terra offers equivalent performance at half the price; Luna focuses on being lightweight and fast. Due to US government requirements, the series shifted from planned open access to a limited preview.

88Z AI editorial1 min read
Flash

DeepSeek and Peking University Open-Source Speculative Decoding Acceleration Framework DSpark

DeepSeek and Peking University have jointly released DSpark, a large model inference acceleration framework now deployed in the DeepSeek-V4 production system. While maintaining total throughput, DSpark significantly boosts single-user generation speed: DeepSeek-V4-Flash sees a 60%–85% increase, and DeepSeek-V4-Pro a 57%–78% rise. It utilizes a semi-autoregressive architecture with confidence-based scheduling verification to balance draft generation efficiency and text coherence, dynamically adju

88Z AI editorial1 min read