#Open-source ecosystem

AI…when it comes to thinking about actions that impact the company that pays them.

AIFireworks AI has made Kimi K3 available for Multi-LoRA serving and training in private preview through Fireworks Serverless Training. The post explains how small LoRA adapters can be trained on K3 and served with live merge or multi-LoRA deployment, and it reports two example tasks, Countdown and Frozen Lake, with reward curves.

AITrilogy released an AI Cybersecurity Playbook built around Kimi K3, served on Fireworks through an OpenAI-compatible API. The playbook pairs deterministic tools for indexing, parsing, and deduplication with Kimi K3 for reachability, impact, and remediation judgment, followed by a separate verification stage.

AISo seeing Microsoft, NVIDIA, and much of the industry sign this letter feels like real progress. But the uncomfortable reality is that we can win the argument while also losing the leaderboard. America invented open source. It's time for us to lead like it.

AIIt lives with scientists, engineers, clinicians, firms. For AI to benefit from distributed knowledge, it must itself be distributed. Agree with Jensen that this is a future worth building.

AIBryan Catanzaro, NVIDIA's account owner, argues the central US AI leadership question is whether AI models will be treated as infrastructure like the internet or electricity. He says open models will be at the heart of this infrastructure, enabling companies from startups to established industry leaders, and making sovereignty possible. He concludes policymakers seeking to keep American AI at the forefront should recognize open models as the critical infrastructure of the AI age.

AIThe essay argues that Western companies increasingly rely on Chinese open-weight models like Qwen and Kimi for post-training, while Western labs cannot lawfully distill from American frontier models. It says Qwen's share of new open-model fine-tunes rose from 1% in January 2024 to 69% by February 2026, citing ATOM's Report. The authors propose controlled teacher access and tighter enforcement against foreign distillation as a domestic alternative.

AI…build and deploy AI in their own way. We're now the biggest institutional contributor on HuggingFace and we expect to continue publishing. It's not charity or a science project - we know that when AI grows, NVIDIA's opportunities also grow. More analysis on the state of open source AI here:

AIAndrew Ng and Rohit Prasad announced OpenWorker, an open-source agent that produces deliverables such as documents, Slack messages, and calendar updates across files and everyday tools. It checks in before consequential actions, runs on Mac with Windows support coming soon, and works with user-supplied API keys for models including GPT 5.6 Sol, Claude Fable, Gemini 3.6, open-weight models, or local Ollama models. Source code is available on GitHub, and the tool requires the user's own API key.

AILeandro von Werra introduces The Stack v3, a dataset of 5T tokens ready for training and 120TB of raw data. He says the dataset is meant to support open code models for cyber defence, and links the download on Hugging Face.

AIMistral announced an expanded global strategic partnership with Microsoft, backed by a multi-billion dollar commitment from Microsoft. The deal accelerates Mistral's AI infrastructure build-out in Europe and makes its open-weight models available to Microsoft customers in Copilot Studio, Azure Foundry, and Azure Local.

AIJetBrains Air now connects to ACP-compatible coding agents, including GitHub Copilot CLI, OpenCode, Pi, and Cline, through the Agent Client Protocol. The release also adds Beta Java and Kotlin navigation and diagnostics powered by the IntelliJ IDEA code engine, local model support through Ollama or LM Studio, and Docker-based agent tasks on Windows.

AISYNAPS-I, a multi-lab Genesis Mission project led by Lawrence Berkeley National Laboratory, uses Meta's open-source SAM 3 and DINOv3 models to segment X-ray and micro-CT scientific imagery. The fine-tuned pipeline, run on 300 A100 GPUs, reduced a grapevine xylem analysis from a month of expert annotation per time step to about 15 minutes. The team can deploy the open models inside secure national lab infrastructure, where research data must remain.

AICongrats @Kimi_Moonshot on K3 and its claimed 2.5x efficiency gain over K2. If open labs keep finding 2.5x a year, compute advantages depreciate fast. The frontier isn't who has the most FLOPs. It's who converts them best. 👇

AIMoonshot AI announced Kimi K3, a native multimodal model with 2.8 trillion parameters and a 1 million token context window. The announcement cites up to 6.3x faster decoding in million-token contexts and about 25% higher training efficiency, and says open weights arrive by July 27, 2026. The author, Soumith Chintala, reposted it with a brief note of congratulations.

AIIt's a good first model but the most impressive part: no hype, grounded writing, interesting details, and well integrated across the whole ecosystem. I wish other big labs would follow that release style!

AIThinking Machines has introduced Inkling, an open-weights model that reasons across text, image, and audio, with full weights made available. The model is available today for fine-tuning on Tinker, and the company also offers an Inkling Playground for trying it out. The author describes Inkling as a foundation model intended for broad capabilities in practical use and customization.

AI…UnslothAI, @modal, @baseten, @lightseekorg, @inferact on VLLM, @radixark on SGLang.

AIIt’s strong on audio in particular, ranking among the strongest open-weights models on VoiceBench, MMAU, and AudioMC.

AI…in advance for permission to invent. Protecting open source AI is now a critical part of ensuring this.

AICognition says its SWE-1.7 model, developed from the open-source Kimi K2.7 Code base, performs as well as or better than leading U.S. frontier models on its new trustworthiness evaluation suite. The suite combines 145 politically sensitive questions, sampled in English and Chinese, with realistic coding scenarios to measure propaganda, censorship, and security behavior. Cognition says SWE-1.7 improves substantially over the base Kimi K2.7 Code model, though the company says the benchmarks are still in development.

AIAlong with MTP, Eagle3 and various ngram-based techniques, the local model performance takes another step up. Special thanks to NVIDIA team and Ruixiang Wang specifically for leading this effort!

AIMistral CEO Arthur Mensch argues enterprises should use open-source models and store their own data to avoid dependence on closed providers that retain customer data. He says companies should build continuous training loops from employee and user interactions, and shrink models to cut deployment costs. Mistral positions its Studio control plane and Forge training platform, deployed on customer infrastructure or via zero-data-retention hosting, as tools for this shift.

AIMistral AI released Leanstral 1.5 on Hugging Face as an open-source code agent model for Lean 4 proof assistant tasks. The model uses 119B total parameters with 6.5B activated per token, a 256k context length, and accepts text and image input. The source gives setup paths through Mistral Vibe and a local vLLM server, with recommended settings of temperature 1.0 and reasoning effort set to high for complex prompts. The model is licensed under Apache 2.0.

AIJim Fan announces ASPIRE, a system where coding agents use multimodal sensory traces from simulation and real robots to run evolutionary search over control programs and add the results to a growing skills library. The post claims up to a roughly 10x reduction in transfer learning tokens for sim-to-real and single-arm to bimanual transfer, and says the full stack will be open-sourced.

AIJim Fan introduces ASPIRE, a system in which coding agents observe multimodal sensory traces and run evolutionary search over control programs to distill skills into a growing library. The post says ASPIRE shares know-how rather than pixels or weights across the sim-to-real gap, reducing transfer learning tokens by up to about 10x. The author also says the full stack will be open-sourced and provides a gallery of 150+ tasks and 90+ skills.

AIHuge thanks to the @vllm_project community for the support and collaboration in bringing efficient long-context OCR to more developers. Check out the recipe and give it a try 👇

AIPaddleFormers 1.2 is released with support for training DeepSeek-V4 and 128K+ long-context training. The update adds Context Parallel, Packing, Document Mask Attention, and the Muon optimizer, plus ultra-fused mHC, CSA, and HCA operators, DeepEP/HybridEP communication, and lossless FP8 training with AutoSubbatch memory balancing. The project is presented as fully open-source and is available on GitHub.