Vana acquired the Memory team; an upgraded version of its application is now available for beta testing.
Related
Kalshi's application for a preliminary injunction in the Southern District of New York was rejected by the judge, and the case will now proceed to the motion for dismissal stage.
According to Foresight News , crypto journalist Eleanor Terrett tweeted that Judge Analisa Torres, who presided over the Ripple case, has once again become the focus of a major legal dispute. In the case of prediction market platform Kalshi in the Southern District of New York (SDNY), Judge Torres rejected Kalshi's application for a preliminary injunction, arguing that New York State betting laws apply to Kalshi's sports event contracts and that this law has not been superseded by the Commodity Exchange Act. This ruling means the case will enter the "motion to dismiss" stage. Legal experts have commented that this represents a major setback for Kalshi's lawsuit in the US financial center and could have a ripple effect on other related cases, particularly those in Connecticut and other SDNY jurisdictions.
Tencent's Hunyuan 3.0 official version open source: Switching to Apache 2.0 removes overseas restrictions, halving the illusion rate.
According to Beating's monitoring, Tencent officially released the final version of its 295B parameter Hybrid Expert Model (MoE) 3.0. The most significant change is the switch to the more permissive Apache 2.0 open-source license, removing the previous regional restrictions that prohibited its use in the EU, UK, and South Korea. When the preview version was released in April, many overseas or multinational teams were deterred from using it due to strict regional restrictions and the 100 million monthly active user limit. The final version removes all compliance obstacles and has undergone significant optimization to address deployment challenges: it incorporates a fast and slow thinking mechanism and adds a 3.8B parameter multi-token prediction (MTP) layer for parallel generation, effectively reducing inference latency. After fine-grained data cleaning and training constraints, the final version reduced the illusion rate of large models from 12.5% to 5.4%, and the error rate in multi-round interactive testing from 17.4% to 7.9%. To address the persistent problem of tool calls easily going astray during agent development, Hunyuan 3.0 has also made targeted improvements. Whether in mainstream scaffolding tools like Cline or CodeBuddy, the accuracy fluctuation of cross-framework tool calls is kept below 4%, resulting in more stable output. The officially released FP8 quantized version also lowers the GPU memory barrier for local deployment and fine-tuning.
Sunscreen was acquired by Fhenix, and the team will continue to advance FHE technology.
PANews reported on July 3 that privacy computing project Sunscreen has announced its acquisition by Fhenix. Sunscreen stated that it has focused on applied cryptography research for the past four years and is optimistic about the Fhenix team's execution capabilities in fully homomorphic encryption (FHE) infrastructure, believing that Fhenix is better suited to bring production-grade FHE to market. Following the acquisition, the Sunscreen team will join Fhenix, and the relevant code will continue to be available under the AGPL open-source license.
The Sui mainnet has been upgraded to support zero-gas transaction simulation and has optimized network stability.
Odaily Odaily reports that Sui has announced mainnet upgrade to version V1.74.1, with the protocol simultaneously upgraded to version 128. Key upgrades include: support for zero-gas stablecoin transfer simulations, facilitating application testing for developers; the mainnet officially adopting a timestamp-based Epoch ending mechanism, further optimizing network efficiency; the addition of Move code static analysis functionality to help developers improve code quality; and optimizations to the validator node connection mechanism and node restart stability, reducing abnormal crashes. Furthermore, this upgrade improves range proof verification, cryptographic performance, and Binary Pools-related protocol mechanisms, providing support for future feature expansions.
B.AI Team Gray-Scale Test Launched: Empowering Enterprises with AI, First Batch of Experience Teams Recruiting
ChainCatcher reports that B.AI has officially announced that its team-level AI collaboration product, "B.AI Team," has entered its first phase of gray-scale testing, and simultaneously launched a global open recruitment for the first batch of experience teams. This version focuses on addressing pain points in enterprise-level AI collaboration, introducing a shared Credits points pool to achieve unified scheduling and cost visualization management of team computing resources; it also provides a flexible multi-level member permission system, supporting fine-grained configuration of model usage strategies based on roles, truly adapting AI capabilities to the collaboration processes of different teams. Starting today, teams interested in gaining an early advantage can submit their requirements through the official application link, along with a brief description of their core use cases; after approval, the B.AI team will immediately assign a dedicated person to connect with them, and will provide priority access rights when the official version is launched, helping teams to be among the first to enjoy the new intelligent collaboration experience.
OKX Flash Earn now features DATA "Trade to Earn Tokens" with a prize pool of 1 million DATA tokens.
According to Foresight News , OKX Flash Earn has launched the DATA "Trade to Earn" feature. From now until 16:00 on July 20th, users who complete designated deposit and spot trading tasks will have the opportunity to participate in winning 1,000,000 DATA rewards. OKX Flash Earn has previously been upgraded to a dual-mode system of earning coins through staking and earning coins through trading. Among them, earning coins through trading supports a reward weighting mechanism and provides a reward calculator. Users can view and participate in related activities through the "Flash Earn" entry at the top of the OKX App's Explore page.