long-context — AI Digest

30 апр DeepSeek V4: official open-source release with Day-0 adaptation for Huawei Ascend DeepSeek models-llm

2 мая xAI completes Grok 4.3 API rollout with 1M context, native video, and ~40% price cut xAI models-llm

2 июн MiniMax Releases M3: Open-Weight Frontier Model with 1M-Token Context and MSA Architecture MiniMax models-llm

8 июн NVIDIA Nemotron 3 Ultra: Open 550B MoE Model Now Available for Agentic Workloads NVIDIA models-llm

10 июн MiniMax M3 Open Weights Released: 1M Context, MoE, Frontier Coding MiniMax models-llm

17 июн Zhipu AI Open-Sources GLM-5.2 Under MIT License with 1M Token Context Zhipu AI models-llm

19 июн Zhipu AI Releases GLM-5.2 Open Weights: 753B MoE with 1M-Token Context under MIT License Zhipu AI / Z.ai models-llm

17 июл Moonshot AI launches Kimi K3, a 2.8T-parameter open-weight model Moonshot AI models-llm

25 авг Apodex 1.1: Scaling Agentic Intelligence for Complex Work Apodex research

11 сен DeepSeek V4.1 Flash goes GA with open MIT weights and V4 Pro retirement DeepSeek models-llm

23 окт MiniMax open-sources M3.1 Flash, a 428B-parameter agentic MoE with 1M context MiniMax models-llm

24 окт Moonshot AI releases Kimi K3.1 Pro, a 3T-class open-weight model with 1M context Moonshot models-llm

28 авг Tencent open-sources Hy4 preview: 770B / 49B-active MoE with Gated DSA + iHC, 1M context, Apache 2.0 Tencent models-llm

15 мая SU-01: Gold-Medal-Level Olympiad Reasoning via Curriculum SFT and Two-Stage RL SU-01 Team research

4 июл A Systematic Analysis of Hybrid Linear Attention: 72-Model Study ByteDance Seed research

14 авг DeepSeek-V4-Pro reaches general availability, prices jump up to 1,100% from Aug 16 DeepSeek models-llm

15 сен Tencent Hunyuan open-sources SAS sparse-attention routers trained end-to-end on Qwen3 Tencent Hunyuan research

23 окт iFLYTEK unveils Spark X2.6 trained entirely on domestic compute, open-sources a 4B model iFLYTEK models-llm

18 мая RoPE Provably Fails at Long Contexts: Locality Bias and Token Consistency Both Break research

14 июн Moonshot AI Releases Kimi K2.7-Code: 1T-Parameter Open-Weight Coding Model with Vision Moonshot AI models-llm

14 июн MiniMax Sparse Attention: 28× Compute Reduction at 1M-Token Context with No Quality Loss MiniMax research

29 июл Kimi K3 Technical Report: Kimi Delta Attention and Stable LatentMoE Architecture Detailed Moonshot AI research

27 авг IBM releases Granite 4.2 reasoning LLMs (3B/8B/30B, Apache 2.0, 512K context) trained on ~15T tokens with GB200 NVL72 GRPO RL IBM models-llm

16 мая MemLens: Benchmark for Multimodal Long-Term Memory in Vision-Language Models NVIDIA research

4 июн Echo-Infinity: Real-Time Infinite Video Generation via Learnable Memory Query research

8 июн GitHub Copilot Gets 1M Token Context Window and Configurable Reasoning Levels GitHub / Microsoft tools

14 июн vLLM Adds Day-0 Support for MiniMax M3 Open Weights with 1M-Context Sparse Attention MiniMax tools

7 июл CompactionRL: Reinforcement Learning with Context Compaction for Long-Horizon Agents Zhipu AI / Tsinghua University research

2 авг Metis: Memory Foundation Model MemTensor, Renmin University, NUS, Shanghai Jiao Tong University, Tongji University research

20 авг Demystifying Agent Skills: Why They Work — Until They Don't UC San Diego (Zhiyuan Jiang, Mengdi Wang, Yijiang Li et al.) research

27 авг Zhipu releases GLM-5.3-Flash, first natively multimodal GLM-5 model Zhipu AI / Z.ai models-llm

27 авг Qwen3.8-Flash-Next released as experimental Qwen4 architecture preview Alibaba (Qwen Team) models-llm

1 сен Qwen team details the Qwen3.8-Next architecture: hybrid attention, n-gram embeddings, Muon Qwen (Alibaba) research

7 сен LatentPress: compressing context into continuous memory tokens a frozen LLM reads directly research

12 сен Ant Group open-sources Ling-3.0-flash-Fin, its first finance-enhanced model Ant Group models-llm

18 сен DeepSeek-V4.1-Flash: Pushing the Limits of KV Cache Compression DeepSeek research

3 июн Do Language Models Need Sleep? Offline Recurrence as Memory Consolidation for Improved Inference Google / CMU research

14 июн Zhipu AI Releases GLM-5.2: 744B MoE with 1M-Token Context and Coding-First Design Zhipu AI models-llm

6 июл AgenticSTS: Bounded-Memory Testbed for Long-Horizon LLM Agents Alaya Studio research

13 июл Jet-Long: Efficient Long-Context Extension with Dynamic Bifocal RoPE MIT / NVIDIA research

17 авг Intern-S2-Mobius: Foundation Model with Decoupled Knowledge and Reasoning Shanghai AI Lab research

25 авг Prime Agent: A Self-Improving RLM Harness Prime Intellect research

25 авг ReWorld: An Interactive World Model with Long-Horizon Memory TongyiLab (Alibaba) research

8 июн SubtleMemory: Benchmark Reveals Agents Systematically Fail Fine-Grained Relational Memory research

10 июн SearchSwarm: Delegation Intelligence for LLM Agents in Long-Horizon Deep Research research

3 июл FlashMorph: Data-Driven Hybrid Attention Layer Placement via Learnable Gates ByteDance Seed research

18 июл LongStraw: Long-Context RL Beyond 2M Tokens under a Fixed GPU Budget Mind Lab research

23 июл vLLM previews production-scale support for Moonshot AI's Kimi K3 ahead of its weight release vLLM / Moonshot AI tools

11 авг Motif 3: Technical Report Motif Technologies research

2 сен llama.cpp Sep 1-2 wave: +4.9% generation from n-gram lookup fix, fused CUDA MoE reduction ggml tools

8 сен Language Models Can Control Their Own Attention research

24 окт vLLM serves 1M-token GLM-5.3 contexts on 16 H200s with PCP, DCP and Hybrid HiSparse vLLM tools