Entity · model
Kimi K2
modelactive
kimi-k2-5bc2b50c·7 events·first seen Aliases: Kimi K2, Kimi-K2, Kimi K2.7
Merged from
Kimi-K2
Co-occurring entities
More like this (12)
Recent events (7)
Moonshot AI releases Kimi K3: 2.8T parameter open-weights MoE with frontier-level performance
Moonshot AI introduces Kimi K3, a 2.8-trillion-parameter Mixture-of-Experts model with 104B activated parameters, native vision, and a 1-million-token context window, released as open weights. The model introduces architectural innovations including Kimi Delta Attention, Attention Residuals, and Stable LatentMoE, achieving approximately 2.5x scaling efficiency improvement over its predecessor Kimi K2. Post-training emphasizes reinforcement learning across general, agentic, and coding domains with multi-level reasoning effort. Evaluations show frontier-level performance across coding, agentic, knowledge, reasoning, and vision tasks, trailing only Claude Fable 5 and GPT-5.6 Sol among evaluated models.
Moonshot AI's Kimi K3 (2.8T-parameter MoE) ranks third on Intelligence Index, first among open-weights models
Moonshot AI released Kimi K3, a 2.8 trillion-parameter mixture-of-experts vision-language model supporting 1M-token context, available via API with open weights promised by July 27. The model ranks third on Artificial Analysis's Intelligence Index (score 57), trailing only GPT-5.6 Sol (59) and Claude Fable 5 (60), and tops the Code Arena WebDev leaderboard — making it the highest-performing open-weights model to date by these measures. Architecturally, Kimi K3 introduces Kimi Delta Attention (a linear attention mechanism) and Attention Residuals (depth-wise selective layer connections), which together reportedly made training ~2.5x more compute-efficient than its predecessor. The article also notes that Alibaba launched Qwen3.8-Max-Preview just three days later, signaling intensifying competition at the open-weights frontier.
Data Points: PrismML fits 27B model on iPhone; Cognition SWE-1.7, Nvidia Audex, Anthropic language-value study
A newsletter digest covers four notable AI developments: PrismML (a Caltech/Khosla spinout) compressed Alibaba's Qwen 27B model to under 4 GB via ternary/binary quantization for on-device iPhone inference; Cognition released SWE-1.7 (trained on Kimi K2.7), jumping from 9.4% to 42.3% on FrontierCode 1.1 Main with novel RL and infrastructure techniques; Nvidia introduced Audex, a 30B unified audio-text transformer trained on 157B audio tokens; and Anthropic published research showing Claude's expressed values shift measurably by language across 309,815 conversations. Each item represents a distinct technical development across on-device inference, coding agents, multimodal models, and model behavior analysis.
Anthropic redeploys Claude Fable 5 after US export controls lifted; details safeguard framework and government collaboration
Anthropic is restoring global access to Claude Fable 5 starting July 1, 2026, after US export controls imposed on June 12 were lifted on June 30. The controls were triggered by an Amazon research report showing a jailbreak that allowed Fable 5 to identify software vulnerabilities and produce exploit code, though Anthropic's own testing confirmed comparable models (including Claude Opus 4.8, GPT-5.5, and Kimi K2.7) could produce the same outputs. Anthropic has deployed an improved safety classifier blocking the reported technique in over 99% of cases, and is co-developing a shared industry jailbreak severity framework with Amazon, Microsoft, Google, and other Glasswing partners. Access to the higher-capability Claude Mythos 5 remains restricted to approved US organizations under the Glasswing program.
Agents-A1: 35B MoE agent matches trillion-parameter models via horizon scaling
Researchers introduce Agents-A1, a 35B Mixture-of-Experts model that claims to match or exceed trillion-parameter models like Kimi-K2 and DeepSeek V4 on long-horizon agentic benchmarks. The approach scales agent trajectory length (averaging 45K tokens) and heterogeneous agent abilities rather than raw parameter count, using a three-stage training recipe including multi-teacher domain-routed distillation. On benchmarks such as SEAL-0, IFBench, HiPhO, and FrontierScience-Olympiad, Agents-A1 achieves leading or competitive results against models with roughly 30x more parameters. The work proposes a practical efficiency path for agentic capability scaling without proportional compute scaling.
Nvidia releases Nemotron 3 Super 120B-A12B open-weights model with hybrid Mamba-2/MoE architecture
Nvidia released Nemotron 3 Super 120B-A12B, an open-weights LLM with a hybrid Mamba-2/transformer/MoE architecture that activates only 12B parameters per token and supports up to 1 million token context. The model claims the fastest inference speed in its size class at 442 tokens/second and leads open-weights models on PinchBench agentic task evaluation, outperforming larger models including Kimi K2.5 (1T parameters). Nvidia is releasing weights, training data, and recipes under a permissive commercial license, and plans a $26B five-year investment in open-weights models — framed partly as a strategic response to Chinese labs building capable open-weights models on non-Nvidia hardware.
Mistral Releases Devstral 2 (123B) and Devstral Small 2 (24B) Coding Models Plus Vibe CLI Agent
Mistral AI has released Devstral 2, a 123B-parameter open-weight coding model scoring 72.2% on SWE-bench Verified, and Devstral Small 2, a 24B model scoring 68.0% on the same benchmark and deployable on consumer hardware. Both models support a 256K context window and are permissively licensed (modified MIT and Apache 2.0 respectively). Mistral also launched Vibe CLI, an open-source terminal-based coding agent powered by Devstral that supports multi-file orchestration, natural language code editing, and IDE integration via Agent Communication Protocol. Devstral 2 is currently free via API with post-free pricing of $0.40/$2.00 per million tokens input/output.