Flash

The news, the context, and what changes.

RSS
Flash

SpaceXAI Renamed to SpaceXSI

Elon Musk's SpaceXAI official account has been officially renamed to "SpaceX Super Intelligence" with the new handle @SpaceXSI. This move confirms Musk's previous statement positioning SpaceX as a super intelligence company, marking a comprehensive brand transformation of its AI business toward super intelligence.

88Z AI editorial1 min read
Flash

HeyGen Launches Proprietary Speech Model HeyGen Voice

HeyGen has released its proprietary speech model, HeyGen Voice, which topped the Artificial Analysis TTS leaderboard. Sharing the same end-to-end tech stack as the Avatar V digital human model, it preserves user tone, stress, emotion, and rhythm to improve reading naturalness. The decoupling of similarity and expressiveness_boost parameters balances fidelity and expressiveness.

88Z AI editorial1 min read
Flash

Infinigence AI Completes 2.3 Billion Yuan Series C Funding

Infinigence AI announced the completion of a 2.3 billion yuan Series C funding round, bringing its total fundraising to 4.3 billion yuan, setting new records in the domestic AI infrastructure sector. Investors include state-owned capital, industrial partners, and prominent firms such as Sequoia China, Tencent, and Baidu. The company focuses on efficiency optimization between chips and models without building data centers or training models. Its platform supports 16 chip types and 37,000 PetaFLOP

88Z AI editorial1 min read
Flash

Musi Intelligence Completes Over 1 Billion Yuan Pre-A Financing

Musi Intelligence announced the completion of its Pre-A financing round, raising over 1 billion yuan. The round was led by Huakong Fund, with participation from Guokai Sci-Tech Innovation, Shanghai Sci-Tech Innovation Fund, and existing shareholders. The funds will be dedicated to researching and developing full-modal real-time interactive intelligence and advancing global commercialization. Incubated by Qiu Xipeng's team, the company aims to extend AI beyond screens into devices like glasses, c

88Z AI editorial1 min read
Flash

ByteDance Merges TraeWork and TraeCode into New TRAE

ByteDance announces the merger of TraeWork and TraeCode into the new TRAE, upgraded to a full-link development platform supporting multi-terminal synchronization. The new platform offers a unified entry point, allowing seamless switching between Agent mode for end-to-end delivery and IDE mode for deep coding, covering desktop, web, and mobile.

88Z AI editorial1 min read
Flash

TypeSafe AI, Developer of Decision Model Jev, Announces $870 Million Funding Round

TypeSafe AI, developer of the decision model Jev, has announced the completion of an $870 million funding round, reaching a valuation of $7.5 billion. Since its launch, the model has gained rapid adoption, being used by approximately one-third of Fortune 500 companies, surpassing one million users, and processing trillions of tokens daily.

88Z AI editorial1 min read
Flash

OpenAI Launches Ultrafast Version of New Model GPT-6.1 Sol

OpenAI has released the Ultrafast version of GPT-6.1 Sol, delivering an 8x speed increase while maintaining intelligence levels close to the flagship Astra model. Designed for real-time interaction scenarios, it is priced at $12 per million input tokens and $60 for output, with access currently limited to specific subscribers and enterprise or education clients.

88Z AI editorial1 min read
Flash

Google Launches General Office Agent Gemini Agent

Google has launched the general office agent Gemini Agent, featuring autonomous planning and cross-application collaboration capabilities. This agent integrates into teams with an independent identity and optimizes task execution through four memory mechanisms, automatically selecting underlying models based on requirements.

88Z AI editorial1 min read
Flash

JetBrains Open-Sources AI Coding Model Mellum2.1

JetBrains has released the open-source model Mellum2.1, featuring a 12B mixture-of-experts architecture. The model makes reinforcement learning central to training, supplements data across multiple domains, and launches millions of sandboxes. It offers capabilities for codebase exploration, file editing, and fault diagnosis, with inference throughput nearly double that of Qwen3.5-9B under high load and a speed increase of approximately 1.6x via multi-token prediction.

88Z AI editorial1 min read
Flash

Arena Completes $200 Million Series B Financing

AI model evaluation platform Arena announced the completion of a $200 million Series B financing round, reaching a valuation of $3.1 billion. The round was led by Lightspeed and Khosla, with participation from a16z, Salesforce, and others. Since its Series A in January, its valuation has nearly doubled within ten months, while annual recurring revenue grew from $30 million to $100 million.

88Z AI editorial1 min read