AllLLMRoboticsStartupsResearchEthicsBusinessHardwareAutonomousMedicineCreative
Topic hub

Open-Source AI

Open models — Llama, Mistral, DeepSeek, Qwen and others — are catching up to closed labs and let you run AI on your own hardware, vendor-free. That shifts the balance of power, the safety debate and the economics of the field. Here is everything on open-source AI: open-weight releases, licenses, local deployment and the race against closed models.

Official site ↗1304 articles
In the glossary:Open-Source AIOpen-Weights ModelLarge Language Model (LLM)Foundation Model
TriAgent: A Committee of Small Models Instead of an Expensive LLM Saves $9.3M a Year
LLM·arXiv cs.CL

TriAgent: A Committee of Small Models Instead of an Expensive LLM Saves $9.3M a Year

Jul 25, 2026·2 min
NLI Hypergraph vs. LLM Judge: A New Way to Audit AI Reasoning Without a Reference Answer
Medicine

NLI Hypergraph vs. LLM Judge: A New Way to Audit AI Reasoning Without a Reference Answer

Jul 25, 2026·2 min
Small Qwen Language Models Ignore Instructions but Look Accurate on Benchmarks
Creative

Small Qwen Language Models Ignore Instructions but Look Accurate on Benchmarks

Jul 25, 2026·2 min
CRA-Bench and CRA-Net: Catching Harm That Builds Up Over the Course of an LLM Conversation
LLM

CRA-Bench and CRA-Net: Catching Harm That Builds Up Over the Course of an LLM Conversation

Jul 25, 2026·2 min
Spectral-LSH: Up to 16× Prompt Compression Without Fine-Tuning to Speed Up LLM Inference
LLM·arXiv cs.AI

Spectral-LSH: Up to 16× Prompt Compression Without Fine-Tuning to Speed Up LLM Inference

The new Spectral-LSH method compresses long prompts 8–16× with no model fine-tuning: it hashes similar tokens into groups in attention's spectral space and…

Jul 25, 2026·2 min
DeepSeek Raised Its Registered Capital 65% in 2026, and Its Founder Says AGI Is the Only Goal
LLM·36Kr (36氪)

DeepSeek Raised Its Registered Capital 65% in 2026, and Its Founder Says AGI Is the Only Goal

DeepSeek's registered capital rose 65% in 2026 after two capital increases, and founder Liang Wenfeng told investors in a closed-door meeting that the path to AGI…

Jul 25, 2026·2 min
LISA: Sparse Attention Speeds Up Long LLM Reasoning by 50% Without Retraining
Creative·arXiv cs.AI

LISA: Sparse Attention Speeds Up Long LLM Reasoning by 50% Without Retraining

The LISA method replaces the standard attention mechanism in language models and speeds up long reasoning by 50% at a 16K-token context, with no training from…

Jul 25, 2026·2 min
Confidential Inference on NVIDIA H100 with Intel TDX: Privacy Costs 17–21% of Speed
Hardware·arXiv cs.AI

Confidential Inference on NVIDIA H100 with Intel TDX: Privacy Costs 17–21% of Speed

Researchers measured the cost of confidential computing for LLM inference: on an NVIDIA H100 under Intel TDX, secure mode cuts throughput by 17–21% and slows time…

Jul 25, 2026·2 min
Top themes
01Laguna S 2.1 vs DeepSeek V4: neolab promises cheaper than Flash and stronger than ProJul 25, 2026·2 min02Tencent Hunyuan's Multimodal AI Chief Hu Han Departs as Team Pivots to World ModelsJul 24, 2026·3 min03Open vs. Closed: GLM-5.2 and DeepSeek V4-Pro Now Trail the Frontier by Just 4–7 MonthsJul 24, 2026·3 min04DeepSeek's $7.4B Round Pushes Asian Startup Funding to a Three-Year HighJul 24, 2026·2 min05Sam Altman's 2022 Letter: OpenAI Considered Open Source to Choke Off CompetitorsJul 24, 2026·2 min
Alibaba unveils Qwen3.8-Max-Preview, which it says trails only Claude Fable 5
LLM·SCMP Tech

Alibaba unveils Qwen3.8-Max-Preview, which it says trails only Claude Fable 5

Jul 24, 2026·2 min
Alibaba's Qwen-Image-3.0: Legible 10-Pixel Text and Full Infographics in a Single Pass
Creative·The Decoder

Alibaba's Qwen-Image-3.0: Legible 10-Pixel Text and Full Infographics in a Single Pass

Jul 24, 2026·2 min
Alibaba Open-Sources Qwen 3.8 Max, a 2.4 Trillion-Parameter Model: Why the US Fears Chinese Models
LLM·Simon Willison

Alibaba Open-Sources Qwen 3.8 Max, a 2.4 Trillion-Parameter Model: Why the US Fears Chinese Models

Jul 24, 2026·2 min
Moonshot AI's Kimi K3: A New 'DeepSeek Moment' and the Chips-vs-Architecture Debate
Hardware·SCMP Tech

Moonshot AI's Kimi K3: A New 'DeepSeek Moment' and the Chips-vs-Architecture Debate

Jul 24, 2026·2 min
Moonshot AI Halts Kimi K3 Subscriptions as Demand Outstrips Compute Capacity
Hardware·SCMP Tech

Moonshot AI Halts Kimi K3 Subscriptions as Demand Outstrips Compute Capacity

Beijing-based Moonshot AI has stopped accepting new Kimi K3 subscribers after demand nearly exhausted the company's compute capacity within 48 hours.

Jul 24, 2026·2 min
Moonshot Releases Kimi K3, the Largest Open Model Yet at 2.8 Trillion Parameters
Creative·Simon Willison

Moonshot Releases Kimi K3, the Largest Open Model Yet at 2.8 Trillion Parameters

Moonshot AI unveiled Kimi K3, a 2.8T-parameter model and the largest open model to date. It beats DeepSeek V4 Pro on size but carries a record price for a Chinese…

Jul 24, 2026·2 min
Microsoft and Mistral Sign Multibillion-Dollar Deal on AI Infrastructure in Europe
LLM·The Decoder

Microsoft and Mistral Sign Multibillion-Dollar Deal on AI Infrastructure in Europe

Microsoft and Mistral AI are expanding their partnership with a multibillion-dollar deal to jointly build data centers and AI compute capacity across Europe.

Jul 24, 2026·2 min
Moonshot AI's Kimi K3: An Escalation in the Open-Weights Race, and What It Means for AI
Business·Interconnects (Nathan Lambert)

Moonshot AI's Kimi K3: An Escalation in the Open-Weights Race, and What It Means for AI

Jul 24, 2026·2 min
Alibaba's Qwen Audio 3.0 TTS Plus Tops the Speech Arena Text-to-Speech Leaderboard
LLM·The Decoder

Alibaba's Qwen Audio 3.0 TTS Plus Tops the Speech Arena Text-to-Speech Leaderboard

Jul 24, 2026·2 min
White House Divided Over How to Respond to Chinese AI Models Like DeepSeek
Hardware·Wired

White House Divided Over How to Respond to Chinese AI Models Like DeepSeek

Jul 24, 2026·2 min
Moonshot AI's Kimi K3: A 2.8-Trillion-Parameter Open-Weight Model Rattles Silicon Valley
Hardware·SCMP Tech

Moonshot AI's Kimi K3: A 2.8-Trillion-Parameter Open-Weight Model Rattles Silicon Valley

Jul 24, 2026·2 min
T-Bank releases T-Search, an open-source agent for multi-step search in Russian
Autonomous·Habr: Т-Банк

T-Bank releases T-Search, an open-source agent for multi-step search in Russian

Jul 23, 2026·2 min
Cisco's Small Open-Source Models vs. GPT-5.5: Finding Vulnerabilities at a Fraction of the Cost
LLM

Cisco's Small Open-Source Models vs. GPT-5.5: Finding Vulnerabilities at a Fraction of the Cost

Jul 23, 2026·2 min
China's Open AI Models vs. OpenAI and Anthropic: What DeepSeek and Qwen Are Offering
Hardware

China's Open AI Models vs. OpenAI and Anthropic: What DeepSeek and Qwen Are Offering

Jul 23, 2026·2 min
Arcee: Chinese AI Models Aren't Inherently Dangerous, Says US Open-Source Lab
LLM

Arcee: Chinese AI Models Aren't Inherently Dangerous, Says US Open-Source Lab

Jul 23, 2026·2 min
Google and Kaggle Open a Free 5-Day AI Agents Course (Agentic AI)
Autonomous·KDnuggets

Google and Kaggle Open a Free 5-Day AI Agents Course (Agentic AI)

Google and Kaggle have opened free access to their 5-day AI agents intensive on Kaggle. There is no fee and no selection process, and you can work through it at…

Jul 23, 2026·2 min
US Treasury vs. Nvidia: Jensen Huang Pushes Back on Sanctions Against Chinese AI
Hardware·TNW

US Treasury vs. Nvidia: Jensen Huang Pushes Back on Sanctions Against Chinese AI

On July 21, 2026, Treasury Secretary Scott Bessent vowed to screen Chinese open-source models for IP theft. Nvidia CEO Jensen Huang says the approach is backwards.

Jul 23, 2026·2 min