Welcome to the "AI Daily" section! This is your guide to exploring the world of artificial intelligence every day. Every day, we present you with the latest content in the AI field, focusing on developers to help you understand technology trends and innovative AI product applications.
New AI products Click to learn more:https://app.aibase.com/zh
1. Google launches Gemini 3.8 Flash TTS: supports natural language customization of voices and 30-second voice replication
Google has launched Gemini 3.8 Flash TTS and Gemini 3.8 Flash-Lite TTS, meeting the needs of developers and creators for personalized character voice dubbing and large-scale audio generation.

【AiBase highlights:】
🎙️ Supports natural language customization of voices, achieving personalized character voice dubbing.
⏱️ 30-second audio samples can efficiently replicate voices, improving the efficiency of speech synthesis.
🌐 Fully supports 100 languages and dialects, providing 2000+ pre-made voices.
2. Tencent launches "Hy Translate" App, supporting 33 languages and offline translation, targeting outbound travel scenarios
Tencent has launched a new translation product called "Hy Translate" App, positioned as a "one-stop translation assistant," targeting scenarios such as outbound tourism and business communication. The app supports bidirectional translation of 33 mainstream languages and five minority languages, and provides multiple translation modes to meet user needs in offline environments. It also supports switching translation styles to improve translation accuracy and naturalness.

【AiBase highlights:】
🌐 Supports bidirectional translation of 33 mainstream languages and five minority languages.
🎙️ Offers multiple modes such as voice translation, photo translation, conversation translation, and offline translation.
💼 Supports switching translation styles to meet different scenario needs.
3. Kimi K3.1 may make its final appearance in October: three levels of reasoning intensity, one million long context, Agent and Swarm fully included in configuration
Kimi K3.1 is expected to be released in October 2026, bringing three levels of reasoning intensity, one million long context, and Agent and Swarm features, marking that the competition of large models has shifted from chat capabilities to actual task processing capabilities.

【AiBase highlights:】
🧠 Kimi K3.1 will support three levels of reasoning intensity, meeting different task needs.
🌐 Supports ultra-long context options, up to 1 million Tokens.
🤖 Introduces Agent intelligent entities and Swarm multi-agent collaboration mode.
4. Huawei officially announced Xiao Yi Work is now available, offering 31 days of free access to 1000 AI points
Huawei officially launched Xiao Yi Work, a new AI work assistant, which is deeply compatible with HarmonyOS devices and offers a 31-day free trial. Xiao Yi Work can complete task decomposition and execution through natural language, covering various office scenarios, while supporting cross-device collaboration and multi-device synchronization, significantly improving work efficiency.

【AiBase highlights:】
💻 Xiao Yi Work supports multi-device collaboration, achieving seamless connection between mobile phones, tablets, and computers.
🚀 Provides natural language task decomposition and execution functions to improve office efficiency.
📊 Supports diversified office scenarios such as document processing, data analysis, and code development.
5. iQiyi's Gong Yu: AI is an added value to content, not a tool for building assembly lines
iQiyi founder Gong Yu emphasized at the iJOY Autumn Enjoyment Conference that AI should be an added value to content, not a tool for industrial production. He mentioned that iQiyi is using AI technology to enhance audiovisual experiences, shorten production cycles, and reduce costs, while working to build a healthy creator ecosystem and explore new forms of works driven by AI, such as interactive videos and mini-games.

【AiBase highlights:】
🎬 AI is an added value to content, not a tool for building assembly lines.
📊 iQiyi is committed to building a healthy content creator ecosystem.
🎮 Exploring new forms of works driven by AI, such as interactive videos and mini-games.
6. Doubao offers all users 30 days of free subscription benefits: standard package free trial until October 17th
Doubao offers all users 30 days of free subscription benefits, with the standard package available for free trial until October 17th. Doubao uses a subscription plus quota model, allowing users to upgrade to paid subscriptions or purchase quota packages to expand usage. This is the latest time Doubao has offered subscription benefits, aiming to lower the payment threshold, cultivate subscription habits, and expand the user base.

【AiBase highlights:】
✅ Doubao offers all users 30 days of free subscription benefits, with the standard package available for free trial until October 17th.
🔄 Doubao uses a subscription plus quota model, allowing users to upgrade to paid subscriptions or purchase quota packages to expand usage.
📈 This is the latest time Doubao has offered subscription benefits, aiming to lower the payment threshold, cultivate subscription habits, and expand the user base.
7. Hy Image3.5preview lands on WorkRally, offering two weeks of free trial for film and television creators
The Tencent Huan Yuan Hy Image3.5preview model has officially launched on WorkRally, providing professional image generation capabilities for film and television creators and offering free access for two weeks. This model has significant improvements in text rendering, realism, and stylistic expression, and supports multi-camera image generation and image re-lighting functions, further promoting the application of AI in film and television production.

【AiBase highlights:】
🎬 The Tencent Huan Yuan Hy Image3.5preview model has joined WorkRally, providing professional image generation capabilities for film and television creators.
🔧 WorkRally platform supports multiple calling paths, covering the creation process from concept design to final scene division.
🌐 OnSolo has also joined Hy Image3.5, providing free trials for overseas short dramas and animations, lowering the creation barrier.
8. Zhiyuan's second 10,000th Yuncheng A3 Ultra is delivered to Changlong, 6,000 units filled in half a year, and embodied robots first put into regular service in theme parks
Zhiyuan successfully achieved mass production of embodied intelligent robots and realized routine application in the cultural tourism field, demonstrating its technical strength and market potential.

【AiBase highlights:】
🤖 The second 10,000th Yuncheng A3 Ultra robot was delivered to Changlong Group, marking the entry of embodied intelligent products into mass production.
🎡 The first large-scale embodied intelligent theme park opened at the Hengqin Changlong Spaceship Park, with over 300 robots operating on a regular basis.
🛠️ Yuncheng A3 Ultra has powerful perception and computing systems, capable of adapting to high-traffic and high-interaction cultural tourism scenarios.
Join Now