Roam has launched an Enterprise eSIM, a customizable management dashboard to empower Web3 teams in cross-border collaboration.

According to GMGN market data, Friend.tech’s FRIEND token has surged 1553% over the past hour, with its current market capitalization standing at $750,000. Today, reports emerged that Huang Licheng, nicknamed "Big Brother Ma Ji", announced plans to acquire Friend.tech for $1 million and take over the FRIEND token.

Royal Bank of Canada (RBC) said it raised Nvidia's price target from $300 to $330.

HIP-3 market deployer Entropy has launched the Anthropic Pre-IPO market on Hyperliquid, with ANTH down 2.22% over the past 24 hours to a current price of $1,974, a 24-hour trading volume of $19.88 million, and contract open interest of $3.44 million.

Crypto figure "Brother Ma Ji" Huang Licheng posted on social media that Friend.tech currently has a market capitalization of less than $300,000. He has made a $1 million acquisition offer to the project’s founder Racer and Paradigm, which would enable a community take-over (CTO) and relaunch of the FRIEND token.

Dongcha Beating AI News Brief: Just one day after the launch of Qwen3.8-Flash, Alibaba Cloud has already cut its pricing. The cost per million input tokens dropped from 1 yuan to 0.8 yuan, while output tokens fell from 3 yuan to 2.7 yuan, marking respective reductions of 20% and 10%. The new Qwen price just undercuts GLM-5.3-Flash, which was released yesterday. GLM-5.3-Flash’s API is priced at 0.8 yuan per input token and 2.8 yuan per output token. Qwen matches GLM’s input price, with output 0.1 yuan cheaper. However, GLM-5.3-Flash offers a 50% discount for its first two weeks post-launch, bringing its current actual price to just 0.4 yuan per input token and 1.4 yuan per output token. During the promotion period, GLM remains nearly half cheaper. Once the discount expires, the prices of these two domestic Flash models will be roughly equal.

Beating AI News Flash: Google has launched Gemini 3.5 Transcribe, its most accurate speech-to-text model to date, and the first dedicated transcription model from Google to feature a Smart mode. The model comes in two versions—real-time and audio recording transcription—and is now open for public beta testing via the Gemini API. In an evaluation by Artificial Analysis, the non-streaming Word Error Rate (WER, lower is better) stands at 2.6%, while the real-time version hits 4.0%. Final transcription latency is 70% shorter than that of its predecessor, Chirp 3. The Smart mode goes beyond literal word-for-word transcription: for example, if a user says “Meeting on Tuesday, no, Wednesday,” it will retain only “Wednesday”; filler words like “um” and “ah” are automatically removed, and spoken content can be organized into paragraphs, lists, dates, and numbers. For verbatim meeting notes, users can switch back to the default literal transcription mode. The model supports over 85 languages, allows mid-transcription language switching, and supports adding custom domain-specific vocabulary. Pricing is approximately $0.005 per minute for audio transcription (equivalent to $5 per 1,000 minutes) and $0.009 per minute for the real-time version. In independent benchmarks, Gemini 3.5 Transcribe is more accurate than OpenAI’s GPT Transcribe (3.3% WER) but lags behind ElevenLabs’ Scribe v2, which has a 2.2% WER and costs roughly $3.67 per 1,000 minutes. The model is already integrated into Android’s Rambler and the macOS version of Gemini, and will soon roll out to Chrome, enabling direct voice-to-text input in web text boxes.