OpenClaw 2.0 Merges 16,000 PRs in One Release
OpenClaw's 2.0 release merges over 16,000 pull requests from 933 contributors in one sweep, reworking setup, browser access, and cloud collaboration.
Open Source News and Analysis
112 verified stories covering Open Source, product updates and industry developments. Page 1 of 2
Tencent Opens Hunyuan Hy4, Narrowly Tops GLM and Kimi in Blind Test
Tencent open-sources Hunyuan Hy4 preview, a 770B MoE model with 6.4% activation and 1M-token context, narrowly beating GLM-5.3 and Kimi K3 in blind tests.
Nvidia Agrees to $12.9B Hugging Face Acquisition
Nvidia has reportedly agreed to acquire Hugging Face for $12.9 billion, nearly doubling last year's rejected $7 billion offer, though the deal remains unsigned and could still fall apart.
SGLang speeds up MiniMax-H3 1.95x with zero quality loss
LMSYS benchmarked MiniMax-H3 on 8x H200 GPUs: SGLang's lossless path beats Diffusers by up to 1.95x, while steeper tiers trade SSIM fidelity for up to 6x speed.
Andrew Ng's OpenWorker Adds Built-In Vulnerability Scanning
Andrew Ng's open-source agent OpenWorker adds a security review role, a fixer-can't-verify rule, and four-tier permissions in its new release.
WeChat's Open-Source Multimodal Embedding Model Tops MMEB-v2
Tencent's WeChat team open-sourced WeMM-Embedding, a multimodal embedding model whose 2B version tops the MMEB-v2 leaderboard, beating rivals four times its size.
Zhipu Open-Sources GLM-5.3-Flash, Cuts Active Params to 18B
Zhipu's open-weight GLM-5.3-Flash halves active parameters to 18B, matches Opus 4.8 on Artificial Analysis, and runs on domestic AI chips at a fraction of the price.
Zhipu Claims Ox Alpha, Usage Tops DeepSeek by 2x
Zhipu confirms Ox Alpha is its new GLM model after anonymous OpenRouter testing, with free-preview usage already more than double DeepSeek's.
Qwen3.8-Flash-Next Launches Tonight, Architecture Jumps Ahead to Qwen4
Alibaba's Qwen3.8-Flash-Next launches tonight on the unreleased Qwen4 architecture, a multimodal MoE with under 5% parameter activation aimed at cutting inference cost.
Meta's Homegrown RDMA Protocol Survives 10% Packet Loss
Meta detailed MetaRoCE, an in-house RDMA transport that skips PFC and still works at 10% packet loss; full spec ships at OCP in October.
Hugging Face Taps Banks to Explore Sale at $13B Valuation
Hugging Face has hired banks to explore a sale near $13 billion, but Bloomberg and TechCrunch say no deal or specific buyer has emerged yet.
Time for Open Models to Catch Up Is Halving Each Generation
SemiAnalysis charts three years of open-vs-closed AI gaps and finds the catch-up window keeps halving, with Kimi K2.6 closing on Opus 4.5 in just 4.8 months.
Ant's Ling-3.0-flash: 2.1x Faster Single-Request Decoding on 4 GPUs
SGLang and Ant Ling Infra cut single-request TPOT for Ling-3.0-flash from 3.33ms to 0.78ms on 4 Blackwell GPUs by removing host-side stalls.
Trillion-Parameter Model Restart Cut From 8.8 Minutes to 32 Seconds
Ant Group and SGLang keep quantized weights resident in GPU memory, cutting Ling-2.6-1T restart time from 8.8 minutes to about half a minute.
Anthropic Embeds Mythos 5 in Security Tools, Grants Open Source $35M
Anthropic folds Claude Mythos 5 into partner security products, funds open-source security with $35M in credits, and widens its Cyber Verification Program.
OpenBMB open-sources 8B model that beats 32B Lean 4 rivals
OpenBMB open-sourced MathForm-8B, an 8B model for autoformalizing math into Lean 4, plus a 367,000-sample verified dataset and evaluation code, outperforming several 32B formalization models.
SenseTime Open-Sources U1.5, 8B Model Outputs Native 4K
SenseTime open-sourced SenseNova-U1.5-8B-MoT with native 4K generation and precise editing, but its parameter count is reported three different ways.
OpenAI Open-Sources Codex's Execution Layer, Cisco Already In
OpenAI open-sourced Codex's execution layer via three integration tiers, with Cisco and a tax-prep firm already using it to build agents.
Liquid AI Cuts Decoding Latency 3.18x With a 300M Draft Model
Liquid AI's LFM2.5-DSpark draft models bring speculative decoding to three on-device models, cutting decoding latency up to 3.18x with identical outputs.
18 AI Models Went Autonomous — Top One Closed 82% of the Gap
Prime Intellect ran 153 fully autonomous research trials across 18 frontier models on a nanoGPT speedrun. The best model closed 82% of the human record gap, but not one of them found a genuinely new method.
Liquid AI revives Q4_0 quantization with 97% accuracy via distillation
Liquid AI's QAD training method restores Q4_0 GGUF checkpoints to 96-97% of BF16 accuracy, preserving the format's speed edge on Arm CPUs and older devices.
Zhipu Opens GLM-5.3 API at $4.40 per Million Output Tokens
Zhipu's GLM-5.3 API prices at $1.40/$4.40 per million tokens and scores 60 on Artificial Analysis's Intelligence Index, ranking 8th of 182 models — but it burns tokens fast.
Qualcomm Open-Sources Mojo Compiler a Month After Modular Deal
Qualcomm has open-sourced Mojo's compiler under Apache 2.0 with the LLVM exception, a month after acquiring Modular — though outside code contributions aren't accepted yet.
DeepSeek Pro adds 384K output
DeepSeek's V4 Pro API now points to the 0813 version, pairing a 1M-token context and 384K output ceiling with a clear price warning.
Alibaba opens Qwen Max-class weights
Qwen3.8-2.4T-A95B has moved from a promise to an open-weight ModelScope repository, giving developers a 4.89TB Max-class model card to test.
Ling tiny model targets local agents
InclusionAI open-sourced Ling-3.0-tiny, a 7.9B-parameter MoE model that activates 1.3B parameters per token and targets local agent workloads.
Meta opens a 30B local agent model
Meta's Muse Glimmer is a 30B open-weight agent model built for local workflows, with Apache 2.0 weights, 24GB and 32GB deployment targets, image input, tool use, and DFlash speculative decoding.
BigBang Lets Models Write Tasks
BigBang-V1 turns synthetic-data production into an agentic loop: models propose, solve, verify, and filter frontier tasks.
Google’s Cyclone AI Adds a Day
Google DeepMind open-sourced WeatherNext Cyclones after Nature results showing roughly one extra day of forecast skill for cyclone track, intensity, and wind structure.
ModelBest Lets Agents Speed Up Industrial Software
ForgeStencil is an OpenBMB system that uses two agents to optimize stencil kernels and deploy verified speedups into real scientific and industrial software.
Inkling Small Closes In on 975B Flagship
Thinking Machines released Inkling-Small with open weights, 276B total parameters, 12B active parameters, multimodal input and benchmark scores close to the 975B Inkling.
MiniMax Opens Its H3 Video Model
MiniMax H3 combines 2K, 15-second video, native stereo audio, RMB 0.8-per-second pricing and a coming open-weight release.
Kimi Valuation Hits $35B
Moonshot AI has reportedly closed more than $3.5 billion in F-round financing at a $35 billion post-money valuation after opening Kimi K3 weights.
Tencent Open-Sources an Inference Speedup Stack
Tencent Hunyuan released AngelSpec, an open-source framework for training speculative-decoding drafters and reducing large-model inference cost.
Robots Get 30,000 Hours of Touch
NeoteAI and Fudan released N0 reports, data and code that put tactile sensing at the center of robot learning.
Kimi K3 Falls Behind in Cyber Tests
UK AISI and US CAISI put Moonshot AI's Kimi K3 through cyber benchmarks, finding strong general-model momentum but a clear gap on exploit chains.
Poolside Releases 118B Open Coding Model
Laguna S 2.1 gives enterprises a self-hosted coding-agent base, trading frontier scores for local control and clearer deployment economics.
MiniCPM puts robot models on device
OpenBMB released MiniCPM-Robot with a 1.5B manipulation model, a 0.9B tracking model, and a deployment path focused on local latency.
Hugging Face discloses AI-agent intrusion
Hugging Face says an autonomous AI-agent campaign exploited dataset-processing paths, while its defenders used local LLM analysis to reconstruct more than 17,000 events.
Kimi K3 arrives with 2.8T parameters
Moonshot AI has launched Kimi K3 with 2.8 trillion parameters, a 1M-token context window and published API pricing, while the open-weight proof still depends on weights and model cards.
SingGuard adds open guardrails for agent actions and multimodal policy checks
SingGuard adds open guardrails for agent actions and multimodal policy checks. The piece reviews the verified facts and why the signal matters beyond one announcement.
Claude-powered Bun rewrite sparks Zig backlash
Claude-powered Bun rewrite sparks Zig backlash. A concise localization of the verified facts and the industry signal behind the story.
Zhipu bets on long-horizon agents after listing
Zhipu bets on long-horizon agents after listing. A concise localization of the verified facts and the industry signal behind the story.
LingBot model predicts before it acts
LingBot model predicts before it acts. A concise localization of the verified facts and the industry signal behind the story.
Ollama raises $65 million for local models
Ollama raises $65 million for local models is reframed for global technology readers, with the key numbers, product claims and open questions kept intact.
LingBot opens a robot vision base model
LingBot opens a robot vision base model is reframed for global technology readers, preserving the key numbers, claims and open questions.
Four Chinese models beat Opus at far lower cost
Four Chinese models beat Opus at far lower cost is reframed for global technology readers, preserving the key numbers, claims and open questions.
GitHub Copilot adds open-weight Kimi K2.7 Code
Model choice, Azure hosting and coding-model price pressure.
Palantir and NVIDIA bring Nemotron into air-gapped AI systems
Sovereign AI, open weights and sensitive government workloads.
Together AI lands $800 million as open-model cloud demand surges
Aramco Ventures led the Series C, valuing the Nvidia-backed open-model infrastructure provider at $8.3 billion.
DeepSeek opens DSpark for faster inference
DeepSeek and Peking University open sourced DSpark, an inference framework that raises generation speed by 60% to 85% and can lift single-GPU throughput as much as 6.6 times.
Meituan opens LongCat-2.0 after Owl Alpha mystery
Meituan identified the high-traffic Owl Alpha model as LongCat-2.0, a 1.6 trillion-parameter MoE model released under the MIT license.
OpenAI’s GPT-5.5-Cyber finds a 23-year-old OpenBSD bug
OpenAI’s GPT-5.5-Cyber finds a 23-year-old OpenBSD bug
OpenAI adds human review to AI-found open-source bugs
Patch the Planet combines Codex Security with Trail of Bits and HackerOne so maintainers receive verified vulnerabilities, severity ratings, patches and tests rather than raw AI reports.
Reflection AI signs a $6.3B compute deal with SpaceX
The young open-source AI lab agreed to rent Nvidia GB300 capacity from SpaceX’s Colossus 2 data center, a deal that can run to 2029 and highlights compute as the scarcest asset.
GLM-5.2 closes in on closed coding models
Z.ai’s open-weight GLM-5.2 scored 74.4 on FrontierSWE, near Claude Opus and above GPT-5.5 in some coding benchmarks, while keeping token prices far lower.
Kimi K2.7-Code puts price before leaderboard glory
Moonshot AI released Kimi K2.7-Code with a HighSpeed mode, strong self-reported tool-use scores and a much lower output price than Claude Opus 4.8.
AI subscriptions strain under heavy users
SemiAnalysis estimates that a fully used $200 ChatGPT Pro plan can represent about $14,000 of API value, exposing the weak economics of all-you-can-use AI pricing.
Microsoft Mirage gives video models spatial memory
Mirage stores latent scene features instead of rendered point clouds, making video world-model generation faster and lighter while improving consistency.
Zhipu opens GLM-5.2 with a 1M-token context
Zhipu's GLM-5.2 pairs a 744B-parameter MoE design with a 1M-token window and a coming MIT-style open-weight release for coding agents.
Kimi K2.7 Code undercuts GPT-5.5 on coding token prices
Moonshot's open MoE model offers far cheaper input and output tokens while improving agent-style coding benchmarks over K2.6.
DiffusionGemma rewrites text generation as denoising
Google’s open 26B MoE model generates text in parallel, trading top quality for local speed above 1,000 tokens per second on H100.
Moonshot open-sources Kimi K2.7-Code for AI coding
Moonshot open-sources Kimi K2.7-Code for AI coding.
Google opens DiffusionGemma for faster text generation
DiffusionGemma applies diffusion-style decoding to text, generating chunks in parallel and showing another path beyond autoregressive language models.
Cohere releases North Mini Code for local coding
Cohere’s 30B-parameter MoE model activates about 3B parameters, supports 256K context and is released under Apache 2.0.
DeepSeek rides the agent token boom
Ramp and OpenRouter data show DeepSeek usage accelerating as long-running coding and data agents make token price a first-order enterprise cost.
Moonshot AI seeks a $30 billion valuation for Kimi
Moonshot AI is already discussing another round that could lift its valuation from $20 billion to $30 billion, with Kimi turning open-source model performance into a financing story.
Reve 2 and Ideogram 4 make layout the next image battleground
Reve 2 and Ideogram 4 make layout the next image battleground. The story explains the announcement, the strategic context and the practical risks for the people or companies affected.
Ideogram opens its 9.3B flagship image model
Ideogram opens its 9.3B flagship image model. The story explains the announcement, the strategic context and the practical risks for the people or companies affected.
Nvidia open-sources Cosmos 3, a world model for robots and self-driving cars
Nvidia has released Cosmos 3, an open-source AI model designed to help robots and autonomous vehicles understand and predict physical world interactions.
MiniMax M3 brings sparse attention, supports million-token context
MiniMax released the open-weight M3 model with 1M-token context, advanced coding ability, and native multimodality, claiming it is the first to combine all three.
Nvidia Open-Sources Nemotron 3 Ultra, But It Still Trails China's Kimi
Nvidia unveiled Nemotron 3 Ultra, a 550-billion-parameter open-source model, but it scored 48 on the Artificial Analysis intelligence index, behind Kimi K2.6's 54.
NVIDIA fuses diffusion and autoregression in new Nemotron model
NVIDIA's Nemotron-Labs-Diffusion introduces a tri-mode decoding approach that boosts throughput up to 6× over Qwen3-8B while maintaining accuracy.
Cohere open-sources 218B model Command A+
Cohere has released its flagship Command A+ model under Apache 2.0, a 218B-parameter MoE architecture that runs on as few as two H100 GPUs, targeting sovereign AI and enterprise deployments.
Hermes Agent passes OpenClaw on OpenRouter
Hermes Agent topped OpenRouter's daily activity chart with 224 billion tokens, ending OpenClaw's long run at No. 1 and pointing to a shift in open-source agent adoption.
Moonshot AI Hits $20 Billion Valuation With Meituan-Led Round
Moonshot AI has raised $2 billion at a $20 billion-plus valuation, putting Kimi at the center of China’s race for open-source large models and a possible Hong Kong listing.
DeepInfra Raises $107M as Nvidia Backs Inference Cloud
DeepInfra's Series B signals that independent inference infrastructure for open models and agent workloads is becoming its own layer.
Mistral Workflows brings AI orchestration to enterprise production
Mistral has opened a public preview of Workflows, a Temporal-based orchestration layer for running enterprise AI processes with approvals, retries and customer-controlled data.
LeRobot flaw exposes robot servers to unauthenticated RCE
CVE-2026-25874 lets anyone who can reach a Hugging Face LeRobot PolicyServer execute code through unsafe pickle deserialization.
Xiaomi open-sources 1.02T-parameter MiMo V2.5 Pro
Xiaomi open-sources 1.02T-parameter MiMo V2.5 Pro. The story explains the announcement, the strategic context and the practical risks for the people or companies affected.
DeepSeek V4 Launches, Market Reaction Muted
DeepSeek released its V4 series on April 24, featuring a 1.6-trillion-parameter Pro model, but markets showed little reaction compared to last year's shock. Analysts say the surprise factor has faded as China's open-source AI output becomes routine.
DeepSeek Cuts V4-Pro Price by 75% Within 72 Hours of Launch
DeepSeek slashed the price of its V4-Pro model by 75% and cut cache-hit pricing across its entire API to one-tenth, just 72 hours after the preview release, intensifying the AI price war.
Qwen3.6-27B Open-Sourced: Beats 397B MoE on SWE-bench Pro
Alibaba's Qwen3.6-27B, a 27-billion-parameter open-source model, outperforms its 397-billion-parameter predecessor on agentic coding benchmarks, driven by a new Thinking Preservation feature.
DeepSeek V4 Launches: 1.6 Trillion Parameters, 8x Cheaper Than GPT-5.5
DeepSeek released V4-Pro and V4-Flash under MIT license, achieving top competitive programming scores while pricing output at $3.48/M — a fraction of GPT-5.5 and Claude Opus 4.7.
Alibaba's 27B Model Matches Its Own 397B Flagship in Benchmarks
Qwen3.6-27B, a dense 27B-parameter model, outperforms the 397B MoE Qwen3.5 on coding benchmarks, thanks to architectural stability and a new Thinking Preservation mechanism. It runs on consumer hardware like a single RTX 4090.
Kimi K2.6 tops HLE open-source leaderboard
Moonshot AI released the full production-grade Kimi K2.6 model, which scored 54.0 on HLE-Full, surpassing closed-source models Claude Opus 4.6 and GPT-5.4. The 1-trillion-parameter MoE model activates only 32B parameters per inference and supports up to 300 parallel sub-agents.
Alibaba Qwen3.6-Max Tops 6 Benchmarks, Shifts to Proprietary
Alibaba released Qwen3.6-Max-Preview, its most powerful model yet, but kept the weights closed. It scored first in six benchmarks and second overall, signaling a strategic shift from open-source flagship to proprietary monetization.
Alibaba open-sources Qwen3.6-35B, achieving 73.4% on SWE-bench
Alibaba's Qwen team has released Qwen3.6-35B-A3B under Apache 2.0, a 35B-parameter MoE model that activates only 3B per token, scoring 73.4% on SWE-bench Verified and outperforming Google's Gemma 4-31B by 21 points.
MiniMax M2.7 Self-Iterates 100 Rounds, Performance Up 30%
MiniMax's M2.7 model used an agent framework to automatically run over 100 training iterations without human intervention, boosting internal benchmarks by 30%. The 230B-parameter sparse MoE model activates only 10B per inference and is now open-source.
NVIDIA open-sources Nemotron 3 family and training data
NVIDIA has released the Nemotron 3 series, including model weights, a 3 trillion token pre-training dataset, reinforcement learning pipelines, and evaluation tools, all under open licenses.
GLM-5.1 Tops SWE-Bench Pro, Trained Entirely on Huawei Ascend
Z.ai (formerly Zhipu AI) released GLM-5.1, a 754B-parameter open-source MoE model that scored 58.4% on SWE-Bench Pro, surpassing GPT-5.4 and Claude Opus 4.6. The model was trained entirely on Huawei Ascend 910B chips with MindSpore, using zero NVIDIA hardware.
NVIDIA Open-Sources 35B Quantum AI Model Ising
NVIDIA has released Ising, the first open-source family of quantum AI models under Apache 2.0. The 35B-parameter Ising Calibration slashes QPU recalibration from days to hours, while Ising Decoding delivers faster and more accurate quantum error correction.
Kimi K2.6 Code Launches, Priced 5x Cheaper Than Claude Sonnet
Moonshot AI has rolled out the K2.6 code preview to all Kimi Code subscribers, offering deeper reasoning chains, improved multi-step agent planning, and more reliable tool calls at a fraction of the cost of competing models.
26-Person Team Builds 40-Billion-Parameter Open-Source Reasoning Model
Arcee AI, a 26-person startup, spent $20 million and 33 days training Trinity-Large-Thinking, a 40-billion-parameter open-source reasoning model that tops agent benchmarks and rivals proprietary models.
DeepSeek V3.2 Switches to Sparse Attention, Halving Inference Costs
DeepSeek quietly released V3.2 with a sparse attention mechanism called DSA, cutting computational complexity from O(L²) to O(Lk) and slashing API prices to $0.28/M input and $0.42/M output, while maintaining comparable benchmark performance.
Mistral Open-Sources 4B Voice Model, Rated More Natural Than ElevenLabs in Human Evaluation
Mistral released Voxtral TTS, its first voice model, on March 26. The 4B-parameter open-weight model is rated by human evaluators as more natural than ElevenLabs Flash v2.5 and on par with ElevenLabs v3. API pricing is $0.016 per thousand characters, and the model can run on your own server.
Mistral Open-Sources MoE Model: 119B Parameters, Only 6.5B Activated
Mistral released Small 4, a single MoE model that merges three previous standalone models into one deployment, supporting text reasoning, image understanding, code generation, function calling, and JSON output with a 256K context window.
Cursor Flagship Product Uses Chinese AI Model
Cursor's new Composer 2 model is built on Moonshot AI's open-source Kimi K2.5, a fact the company initially omitted from its launch materials. The revelation has sparked debate about transparency in AI development.
Google switches Gemma 4 to Apache 2.0 license, releases four models
Google DeepMind released Gemma 4 on April 2, switching to the Apache 2.0 license and ranking third on the Arena AI open-source leaderboard with the 31B model.
Meta's Next Flagship Models Avocado and Mango Will Be Partially Closed-Source
Meta is developing two new flagship AI models, Avocado and Mango, but unlike the fully open Llama series, these will only be partially open-sourced, with key capabilities like cybersecurity code generation withheld.