open-weight-models
A task factory ran fifteen rounds and broke the model grading it News
A new paper builds harder and harder terminal tasks by recursively rewriting accepted ones, and across fifteen rounds a fixed frontier solver's success rate fell from 90 percent to 2.5 percent, with the authors reporting no ceiling in sight.
DeepSeek warns of a significant API price rise, five days after being called 100 times cheaper News
DeepSeek added a footnote to its official pricing page warning that it plans to raise API prices significantly in the near future with no figure and no date attached, five days after an independent benchmark study priced its model at roughly 100 times less per task than Western frontier models.
Ant published Ling-3.0-flash's weights under plain MIT, with no rider attached News
Ant Group's InclusionAI lab published the full weights for its 124-billion-parameter Ling-3.0-flash model on Hugging Face this week under an unmodified MIT license, with no acceptable-use policy, revenue threshold, or branding requirement anywhere in the release.
Model merging: combining two fine-tuned models by averaging their weights Lesson
Model merging combines two or more models that share a common ancestor by arithmetic on their weights - averaging them, adding their differences, or interpolating between them - producing a single model with both sets of skills and no additional training.
llama.cpp shipped DSpark for DeepSeek V4 Flash, and almost everyone called it the wrong name News
llama.cpp release b10228 merged speculative-decoding support for DeepSeek V4 Flash, but the new 0731 checkpoint embeds DSpark and ships no MTP at all, so the widely repeated "MTP support landed" advice points users at a head their model does not contain.
Kimi K3 runs in 8 gigabytes of RAM, at 33 seconds per token News
A hand-written C engine generates text with Moonshot's 2.8-trillion-parameter Kimi K3 using a peak of 8.24 gigabytes of RAM and no GPU, by reading the model's four-bit experts directly off disk, at a rate of roughly one token every 33 seconds.
Beijing says U.S. firms distilled Chinese models, and names none of them News
China's Ministry of Commerce said in a written statement on 27 July that many U.S. AI companies had distilled Chinese models during research and training, identifying no company, no model and no evidence, mirroring a U.S. accusation five days earlier that named two companies but published no logs either.
Three separate tricks dropped the hardware floor for local AI in one day News
A 26-billion-parameter Gemma model ran on an iPhone by streaming expert weights from storage, AMD shipped a sparse mixture-of-experts model activating 2.8 billion parameters per token, and a llama.cpp fork began saving conversation caches to disk - three unrelated attacks on three different bottlenecks.
The open-weights industry letter grew from 25 names to 35 - and OpenAI is on it News
A cross-industry statement titled Open Weights and American AI Leadership now lists 35 signatories on its live Microsoft-hosted page, including OpenAI, Nous Research, GitHub and Cisco, contradicting the widely shared claim that OpenAI declined to sign.
Poolside's Laguna S 2.1 shipped with a broken chat template - and the fixes explain the reviews News
Days after releasing its open coding model, Poolside has been repairing it in public: the base chat template shipped with reasoning disabled by default and a 32,768-token generation cap, and its own quantised builds needed re-releases to fix tool calls and thinking.
AREX is a 4B research agent that re-runs its own research when it doubts the answer News
Beijing Academy of AI released AREX, a deep-research agent whose outer loop checks a provisional answer against the original question's constraints and decides whether to accept it, refine it, or restart the search - with a 4-billion-parameter version released under Apache 2.0.
Safety institutes measure Kimi K3's hacking ability: better than any open rival, nowhere near the top News
The UK and US AI safety institutes found Kimi K3 scored 32% on an exploit-development benchmark versus 24% for the previous open leader, but reached working code execution in zero of 41 attempts where leading closed models average about half.
Black Forest Labs launches FLUX 3 -- image, video, audio, and a robot that never renders the video News
Black Forest Labs announced FLUX 3 on July 23 as a unified image, video and audio model with 20-second video and native audio, available only in early access, with open weights deferred to a later Dev release.
A Huawei-chip training report shows what leaving CUDA actually costs News
SLAI's technical report documents full-parameter post-training of a DeepSeek-V4 model on Huawei Ascend hardware, and the work list -- rebuilt collectives, converted checkpoints, hand-written kernels -- is the real measure of chip independence.
White House Says Moonshot Distilled Anthropic's Fable to Build Kimi K3 News
OSTP Director Michael Kratsios said the US government has information that Moonshot AI distilled Anthropic's Fable model to build Kimi K3, but no supporting evidence has been made public.
Arcee Announces Genesis-Science-1, an American Open-Weight Model for Science News
Arcee announced Genesis-Science-1, a planned trillion-parameter-class American open-weight scientific model built with the Department of Energy and national labs, with weights promised later this year.
US floats sanctions over AI 'distillation' as Anthropic details 16 million scraped chats News
The Treasury secretary suggested the US could sanction Chinese AI labs over model 'theft' while Anthropic and OpenAI allege large-scale unauthorized scraping of their models' outputs, but the verified record shows provider allegations and a proposed sanctions bill, not enacted policy or any proof that model weights were copied.
TimeLens2 teaches video AI to point to the exact seconds that answer a question News
Researchers released TimeLens2, an open-weight video model fine-tuned to answer a text query by returning the exact timestamp intervals in a video that contain the evidence, using a new distance-sensitive reward and a carefully curated dataset, with small versions reported to beat much larger open models on temporal grounding.
Poolside's Laguna S 2.1 is a small open coding agent with big benchmark claims News
Poolside released Laguna S 2.1, a public-weight coding model with an unusually low 8 billion active parameters that runs locally on a single high-end machine, but its claims of beating DeepSeek V4 Pro come from the company's own benchmark table and one early hands-on tester found it fabricates facts when evidence runs out.
Nanbeige4.2-3B reuses one 22-layer stack twice to punch above its size News
A Chinese lab released Nanbeige4.2-3B, a small open-weight model that runs its 22 transformer layers twice in sequence to get 44 layers of depth from one set of weights, posting benchmark numbers rivaling models three times its size, though the results are vendor-reported and a widely repeated 'beats 4x its size' claim does not survive clean accounting.
Model extraction attacks: stealing an AI through its own API Lesson
A model extraction attack tries to copy a machine-learning model you can only query, not download, by sending it many inputs and learning from its outputs. Depending on the goal, an attacker can clone the model's behavior, recover pieces of its internals, or reconstruct a rival model cheaply, which is exactly the fear driving today's AI 'distillation' disputes.
Looped transformers: reusing the same layers to think deeper Lesson
A looped transformer runs its input through the same stack of layers more than once, so it gets the processing depth of a deep network while storing the weights of a shallow one. It trades extra compute for a smaller memory footprint, and it is the trick behind small models that punch above their parameter count.
A Chinese open-weight model is now shipping inside GitHub Copilot News
GitHub has made Moonshot AI's Kimi K2.7 Code generally available as a selectable model in GitHub Copilot, hosting the Chinese-developed open weights on US Azure infrastructure so prompts never reach Moonshot, while a viral claim that Microsoft is secretly testing the newer Kimi K3 remains unconfirmed.
Basalt Labs' 'Best AI Model' Claim Collapses: Its Own Repo Admits Monolith-1.0 Was a Relabeled 7B Model News
Basalt Labs claimed its Monolith-1.0 model scored 99.4% on Humanity's Last Exam and was a 1.57-trillion-parameter system, but its own Hugging Face model card now says the publicly released model was an inflated version of the much smaller Qwen 2.5 7B Instruct, and the weights have been pulled.
A startup router is giving away 100 million tokens of Kimi, MiniMax and GLM News
API aggregator Dahl Inference is handing out 100 million free tokens across top open-weight Chinese models like Kimi K2.6 and MiniMax M2.7 - not a price cut from the labs themselves, but a router burning money to win users amid a glut of cheap compute.
China's GLM-5.2 Ships as the Top Open-Weight Model, Under MIT License News
Z.ai released GLM-5.2, a 753-billion-parameter model, as open weights under an MIT license, and an independent index ranks it the strongest open-weight model available, close behind the leading closed models at a fraction of the price.
A New Campaign Argues You Have a Right to Run AI on Your Own Computer News
A grassroots advocacy site, Right to Local Intelligence, is campaigning against proposed state laws it says could require a license just to download and run open AI models, framing local AI as the next personal computer.
Z.ai's GLM-5.2 claims the open-weight coding crown with a usable 1M-token context News
Z.ai released GLM-5.2, an agentic coding model with a reliable one-million-token context and top open-source scores on long-horizon software benchmarks, with an MIT-licensed weight release promised within weeks.
Ollama nearly doubles Gemma's speed on Macs by guessing ahead News
A free local-AI tool now runs Google's Gemma model far faster on Apple computers using a trick where a small model drafts words and the big one checks them in bulk.
Mistral releases a lean, open model built for formal math proofs News
Leanstral 1.5 is a free, open model specialized for writing machine-checked mathematical proofs, using a design that keeps only a small slice of itself active at a time.
A 35-billion-parameter agent that punches like a trillion-parameter model News
Shanghai AI Lab argues you can reach giant-model performance on long tasks not by adding parameters, but by training on much longer chains of real work.
An open model from China beat Claude on a security test -- at a sixth of the cost News
Semgrep ran GLM 5.2 against Claude on a narrow vulnerability-finding task and the free, open-weight model came out ahead for far less money.
A model that rivals the frontier now squeezes onto a single high-end desktop News
Aggressive compression shrinks GLM 5.2 by more than 80 percent while keeping most of its accuracy, putting a near-frontier model within reach of local hardware.
Quantization: Shrinking AI Models to Run on Modest Hardware Lesson
Storing a model's numbers with less precision - 8, 4, or even fewer bits instead of 16 - makes it dramatically smaller and faster, often with almost no loss in quality. It's why big models can run on a laptop or a single GPU.
Frontier AI is getting more expensive while open models keep getting cheaper News
Closed frontier models are raising prices and tightening access just as Chinese open-weight models slash theirs, a structural reversal with big consequences for who builds with AI.
The Model Ban Is Quietly Redrawing the AI Map News
Two weeks after the US pulled its top models off the market, a Chinese open model sits atop the global download charts and the community is busy rebuilding the banned capability in the open.
AI Agents Are Learning to Build the Worlds They Train In News
Three new open research projects point the same way: instead of only learning what to do, agents are learning to simulate the environment itself, so they can practice in their own imagination.
A Free Model That Splits Your Work Across 300 Helpers News
Moonshot AI's Kimi K2.6 is a frontier-grade model anyone can download, and its headline trick is fanning a single job out to hundreds of helpers working in parallel.
Qwen-AgentWorld Tool
Alibaba's open language world model that simulates agent environments -- browser, terminal, phone, coding workspace and more -- so other agents can be trained inside the simulation. Released with open weights and code in two sizes.
Poolside Laguna S 2.1 (GGUF) Tool
Open-weight 118B mixture-of-experts coding agent activating about 8B parameters per token, under the permissive OpenMDW-1.1 licence, in GGUF plus FP8, NVFP4 and INT4 builds. Use the current re-released Q4/Q8 files - the initial ones shipped with a broken chat template.
MiniMax-M3 Tool
A natively multimodal open model trained on text, image, and video from the first step, with a million-token context and a sparse-attention design built for speed; downloadable for self-hosting and also offered through MiniMax's own API and agent platform.
Kimi K2.6 weights (Hugging Face) Tool
The actual Kimi K2.6 model weights, published under a modified-MIT license for anyone to download, run, and build on; large enough that full-strength use needs a multi-GPU node.
Kimi (Kimi K2.6) Tool
Moonshot AI's web assistant and agent, running the open-weight Kimi K2.6 model; free to use in the browser for chat and long-horizon agent tasks, with the weights also downloadable for self-hosting.
GLM 5.2 (GGUF, runnable locally) Tool
Zhipu AI's open, MIT-licensed mixture-of-experts model with a roughly million-token context, now packaged as ready-to-run quantized files you can host on your own machine. Strong on agent and coding workflows; this week it beat Claude on a narrow security benchmark at a fraction of the cost.
DeepSeek V4 Flash 0731 Tool
The current V4 Flash checkpoint, with weights, the DSpark draft head embedded, and the encoder file that reveals the reasoning-effort labels are prompt prefixes rather than a compute dial. The card also specifies the intended FP8 key-value cache and FP4 indexer cache serving recipe.
Dahl Inference Tool
Third-party inference router reselling top open-weight models (Kimi K2.6, MiniMax M2.7, GLM 5.2) at low per-token prices, currently running a 100M-free-token promotion.
AREX-Turbo Tool
Apache-2.0 4-billion-parameter deep-research agent from BAAI that audits its own provisional answers against the question's constraints and re-runs research when confidence is low. The public quick-start exposes search and page-visit tools but not the paper's full outer control loop.