China AI

AIXiaomi has launched MiMo-V2-TTS, a speech synthesis model that lets users describe the desired voice style in plain language.

Why it matters: The source gives concrete controls for emotion, dialect, singing, and non-verbal sounds, showing how a voice model can be directed through plain-language style prompts.

AIApple has released apple/SimpleSD-30B-instruct, a research checkpoint built on Qwen that uses Simple Self-Distillation to improve code generation without rewards, verifiers, or teacher models.

AIApple has published SimpleSD-4B-thinking on Hugging Face, a research checkpoint built on Qwen that improves code generation through Simple Self-Distillation without rewards, verifiers, teacher models, or reinforcement learning.

AIApple has released SimpleSD-4B-instruct on Hugging Face, a research checkpoint fine-tuned from Qwen3-4B-Instruct-2507 on its own sampled outputs to improve code generation.

AITri Dao announced Mamba-3, which he described as the most powerful linear sequence model to date, as hybrid architectures increasingly rely on strong linear models.

AIBAAI has released RoboBrain-Dex, a dexterous manipulation model for embodied intelligence trained on large-scale, diverse human egocentric data rather than massive robot teleoperation datasets.

AIFun-CineForge, from FunAudioLLM, is an open-source toolkit with an end-to-end dataset pipeline and an MLLM-based model for zero-shot movie dubbing across diverse cinematic scenes.

AIChina's Ministry of Industry and Information Technology launches the Industrial Data Foundation Action, a pilot program to build high-quality industry datasets for AI. Consortia led by key industrial firms, platform institutions, advanced manufacturing clusters and SME digital transformation pilot cities will develop data platforms, standards and datasets, with the goal of building a batch of such datasets by the end of 2026.

AIMiniMax describes Forge, its internal reinforcement learning framework for training real-world agents, which was used during the development of MiniMax M2.5.

Why it matters: The post details how the Forge framework balances throughput, stability, and agent flexibility, with concrete scheduling and prefix-merging methods for training agent RL at scale.

AIMiniMax has released M2.5, which it says reaches 80.2% on SWE-Bench Verified and 76.3% on BrowseComp with context management.

Why it matters: The source gives benchmark scores against Claude and GPT models plus per-task token and runtime figures, so readers can weigh the cost-speed tradeoff directly.

AIZ.ai's release notes list GLM-5.3-Flash, a hybrid-architecture model with 320B total parameters and 18B activated, and GLM-5.3, which the company says achieves a 50% gain over GLM-5.2 on Z.ai Code Bench.

AIZ.ai launches GLM-5, scaling from 355B to 744B total parameters with 40B active and pre-training data from 23T to 28.5T tokens.

Why it matters: The source gives concrete scale, data, and benchmark comparisons against named frontier models, showing where GLM-5 sits among open-source and proprietary systems.

AIZ.ai has launched GLM-OCR, a compact, high-performance optical character recognition model built on its self-developed CogViT and GLM-0.5B encoder-decoder architecture.

AIThe Beijing Academy of Artificial Intelligence (BAAI) published its Emu3 multimodal large model research in Nature, which the post describes as the first large-model achievement led by a Chinese research institution in that journal.

AIZ.ai has released GLM-OCR, a 0.9B-parameter multimodal OCR model for complex document understanding, under the MIT License.

Why it matters: The page gives benchmark scores, a 0.9B parameter size, and supported serving frameworks, which help readers weigh OCR deployment options against heavier alternatives.

AIBAAI released RoboBrain 2.5, an embodied AI model that turns 2D scene understanding into actionable 3D trajectories and provides dense temporal value estimates for real-time progress feedback on long-horizon tasks.

AIBAAI's RoboCOIN is a bimanual robot dataset with more than 180,000 trajectories across 421 tasks, collected from 15 robot platforms in 16 real-world scenarios.

AIZ.ai has released GLM-4.7-Flash, a 30B-A3B MoE model that it positions as the strongest model in the 30B class. The model reports SWE-bench Verified 59.2 and τ²-Bench 79.5, and supports local deployment through vLLM and SGLang.

Why it matters: The source lists benchmark scores against Qwen3-30B-A3B-Thinking-2507 and GPT-OSS-20B, letting readers compare the 30B-class MoE model directly with its named rivals.

AIZ.ai has launched GLM-4.7-Flash, a lightweight, efficient model positioned as the free-tier version of GLM-4.7.

AIZ.ai has launched GLM-Image, an image generation model built on a multimodal architecture that combines autoregressive semantic understanding with diffusion-based decoding.

AITsinghua AIR and BAAI's DrugCLIP screened 10,000 proteins against 500 million molecules, identifying over 2 million drug candidates.

AIMoonshot AI released Kimi K2.5, an open-source native multimodal agentic model built by continual pretraining on about 15 trillion mixed visual and text tokens.

Why it matters: The model card gives a full benchmark table against GPT-5.2, Claude 4.5 Opus, and Gemini 3 Pro, useful for comparing open multimodal agent models.

AIFunAudioLLM has released Fun-Audio-Chat-8B, a roughly 8B-parameter large audio language model for natural, low-latency voice interaction, under Apache 2.0.

AIXiaomi's MiMo model was evaluated against two comparison models on creative writing and artistic perception tasks, with nine evaluators grading outputs on a 1-to-5 scale.

AIMiniMax has released MiniMax-M2.1 model weights on Hugging Face, with API access on the MiniMax Open Platform and the MiniMax Agent product.

Why it matters: The release pairs open weights with a broad benchmark table against Claude and GPT models, letting readers compare coding and agent claims directly.

AIMiniMax released the technical report and pretrained weights for VTP-Large-f16d64, a visual tokenizer that jointly optimizes contrastive, self-supervised, and reconstruction losses.

AIXiaomi released and open-sourced MiMo-V2-Flash, a Mixture-of-Experts model with 309B total and 15B active parameters, under the MIT license.

Why it matters: The post gives architecture, speculative decoding speedup, and pricing figures, which help readers judge how the efficiency claims are achieved and what they cost.

AIAlibaba's FunAudioLLM released Fun-ASR-MLT-Nano-2512, an 800M-parameter multilingual speech recognition checkpoint on Hugging Face that supports 31 languages, with emphasis on East and Southeast Asian languages.

AITongyi Lab has released Fun-ASR-Nano-2512, an end-to-end speech recognition large model trained on tens of millions of hours of real speech, supporting low-latency real-time transcription across 31 languages.

AIDeepSeek has upgraded both deepseek-chat and deepseek-reasoner to DeepSeek-V3.2. deepseek-chat maps to the model's non-thinking mode, while deepseek-reasoner maps to its thinking mode.

AIDeepSeek-V3.2-Speciale is available through a temporary endpoint at the same pricing as V3.2, but without tool-call support. The endpoint remains available until Dec 15, 2025, 15:59 (UTC).

AIMoonshot AI released Kimi K2 Thinking, an open-source thinking model that interleaves step-by-step reasoning with tool calls across 200 to 300 sequential invocations.

Why it matters: The model card gives benchmark tables, quantization details, and deployment settings, letting readers compare Kimi K2 Thinking against GPT-5 and other models on specific tasks.

AIMoonshot AI released Kimi Linear, a hybrid linear attention architecture with 48B total and 3B activated parameters and a 1M-token context length, on Hugging Face.

Why it matters: The model card gives concrete long-context speed and memory figures for a hybrid attention design, useful for judging whether linear attention can replace full attention in practice.

AIMoonshot AI has released Kimi-Linear-Base and Kimi-Linear-Instruct, both 48B total and 3B activated parameters with a 1M context length, on Hugging Face.

Why it matters: The model card gives concrete throughput and KV cache figures for a hybrid attention design, which helps readers weigh its long-context tradeoffs against full attention.

AIA ByteDance research team won the Best Paper Award at the Association for Computational Linguistics (ACL) 2021 for VOLT, a simple and efficient machine translation solution. The team has open-sourced VOLT for developers to use.

AIByteDance's R&D team created the technology behind TikTok's "Cartoonify" sticker in just 48 hours during its annual Hackathon. The augmented reality effect makes physical objects on screen dance, cry, sing, and perform other actions.