Search AI Products and News

OpenAI engineer Tibo announced that the 1 million Token context window of the GPT-5.6 Sol model in Codex has been expanded from being limited to API keys to being available for ChatGPT accounts. Subscribers can manually activate the million-token window, solving the limitation of being forced to truncate historical records when processing large projects.

On Aug 14, Guangdong's first token-economy financial product, Token Loan, launched in Haizhu, Guangzhou for computing/AI SMEs. BOC Guangzhou Branch's Computing Power Token Loan bases quota on contract amount and token consumption, with credit, receivables pledge, order financing, plus guarantees/mortgages to expand.....

Anthropic to launch Sonnet 5.5 (codenamed Fennec) next month, featuring 2M token context, fast low-latency reasoning, improved multi-step planning, and enhanced browser/terminal tool use. Performance rivals flagship Claude Fable 5 but retains Sonnet pricing for best value.....

Wuwen Xinqiong and MiniMax signed a strategic partnership focusing on four areas: jointly optimizing model inference efficiency, building large model token services and AI infrastructure, co-innovating application paradigms, and expanding industry intelligent scenarios. They complement each other: Wuwen excels in computing power scheduling and inference infrastructure, while MiniMax has self-developed large models and massive application traffic.....

AGI Bar in Beijing has launched a 'Token Free Drinking' service, allowing customers to freely and unlimitedly access the DeepSeek V4 Flash model API after connecting to the store's Wi-Fi. The bar is equipped with an NVIDIA DGX Spark AI workstation for local inference, and customers can use the Base URL and API Key to use it in code tools.

The Token Operation Center in Jiaxing, Yangtze River Delta, has launched, offering integrated AI implementation services. The platform has launched a website, providing unified API, Token billing, cost settlement, policy discounts, and security audit. Enterprises can flexibly call models such as DeepSeek and Ali, promoting the standardization of AI computing power and model scheduling in the Yangtze River Delta.

Tencent open-sources the unified training framework AngelSpec, which reduces the cost of autoregressive decoding for large models and improves inference throughput through collaborative schemes and heterogeneous workload adaptation. It adopts differentiated training strategies for different text characteristics in scenarios, such as using a lightweight and stable approach for high-entropy multi-turn dialogues, achieving efficient inference in real-world scenarios.

Alipay upgrades its AI developer incentive program, offering up to 5,660 yuan in token subsidies and lowering barriers. Individual developers integrating AI pay-as-you-go, AI web, or mobile payment can benefit from now until August 21.....

Zhou Hongyi launched NanoWork, an AI-powered enterprise agent platform for bosses, entrepreneurs, solo companies, and employees. It promotes agent collaboration for super organizations. He urges bosses to lead AI adoption, offering 100M tokens each for initial users and hands-on support to help 1,000 small businesses implement AI.....

OpenRouter launches an audio transcription endpoint with a unified API key. Developers just need to send Base64 audio to directly get transcribed text, without needing to integrate Whisper or third-party SDKs separately, completely eliminating the sense of separation between chatting and transcription.

Alibaba released Qwen-Image-3.0, supporting up to 4.5K-token prompts to generate complex content like formulas, geometry, logic, and multi-layer UI. It natively renders 12 languages and 20+ fonts, with 4.5x longer text vs. prior version, advancing commercial creative tasks such as film storyboards and knowledge diagrams.....

Google is developing Frozen v2, an AI chip that hardwires Gemini's architecture into silicon to reduce compute and data movement, boosting inference efficiency. It promises higher energy efficiency and more cost-effective large model execution than the latest general-purpose TPUs.....

Alibaba officially launches the Token Plan personal version and opens up early access to the 2.4 trillion parameter large model Qwen3.8-Max-Preview. This model performs outstandingly in coding and office scenarios. The official version will be open-sourced. Users can obtain model usage rights through Token Plan. Alibaba also launches a limited-time promotion: Lite plan for 39 yuan/month, Standard plan for 139 yuan/month, Pro plan for 499 yuan/month; Team versions offer standard and advanced options.

The main battlefield of large model services is shifting toward retrieval-augmented question answering, multi-document summarization, and long-range agents. The request prompt is composed of dozens to hundreds of semantically independent segments (retrieved documents, skills explanations, memory, historical rounds), forming ultra-long context with tens of thousands to hundreds of thousands of tokens. The pre-filling stage dominates the computing costs, becoming the most prominent cost source for service providers, and triggering more challenging problems.

Multiple banks have recently launched AI-themed credit and debit cards, integrating intelligent benefits such as large model tokens and AI membership into card products to attract young and tech-savvy customers. On July 15, China UnionPay announced the collaboration with multiple banks to create intelligent entity-themed credit cards, further advancing the integration of AI benefits with financial consumption scenarios. Since June, more than a dozen banks including China Merchants Bank, Agricultural Bank of China, Ping An, and SPDB have partnered with China UnionPay, American Express, and platforms like Kimi to take the lead in these initiatives.

At the G7 summit, Mistral AI CEO Arthur Mensch highlighted France's advantage in AI due to abundant low-cost nuclear power, noting France exported 92 TWh of electricity last year. He stressed electricity is key to AI development.....

Elon Musk's SpaceXAI released the new generation large model Grok 4.5, positioned as a versatile main force, capable of handling complex tasks such as code writing, application development, document processing, and knowledge writing, aiming to consolidate its technological advantage in the general AI field and intensify industry competition.

Tencent Cloud Development Enhances the WeChat AI Mini Program Growth Plan: Offering Free 1 Billion Token Model Quota to Developers, AI Image Generation Quota Rises from 10,000 to 100,000, Significantly Reducing Usage Costs. Core Models Are Also Upgraded.

High API costs in AI coding are a pain point. Pxpipe, a local proxy, converts large context into images to bypass costly text token billing, greatly reducing Claude Code expenses. It precisely identifies and transforms verbose text before requests, achieving cost savings by using images.....

Google has significantly increased free quotas for its Gemini API, raising the per-minute token limit for Gemini 2.5 Flash and Flash-Lite models to 1 million for some accounts. This move, already implemented, reduces developer costs and highlights ongoing compute competition.....