SevenTnewS

open-weight models

11 published articles

LLMs & Models5 min read

Open-weight AI

Xiaohongshu's dots3-note: a 280B open MoE that only activates 16B

Xiaohongshu's dots studio has released dots3-note preview, its first open-weight model: a multimodal MoE with 280B parameters, 16B active, and a 512K context window. The sparse design targets low serving cost on one 8-GPU node, but the card has not published benchmark numbers yet.

2026-08-20

Qwen / Alibaba4 min read

Enterprise AI

Forget the model race: Alibaba is automating security operations

Qwen3.8-Max grabbed the headlines, but Alibaba's real move is quieter: AI agents that run security operations inside its cloud console. We break down the SecOps Agent, the Qwen-powered fraud forensics, and the lock-in strategy behind the scores.

2026-08-06

LLMs & Models5 min read

Artificial Intelligence

LFM2.5-Encoders make the small-model case: 3.7× faster than ModernBERT on CPU

Liquid AI's open-weight LFM2.5-Encoders make the case that production NLP belongs on small models. A 230M encoder beats ModernBERT-base on benchmarks and scans full documents in about 28 seconds on a laptop CPU, roughly 3.7× faster.

2026-08-05

AIFeatured2 min read

AI Policy

41 Rivals Signed the Same Open-Weight Letter. That's Not Consensus, It's Nerves

Forty-one rival companies signed the same letter arguing open-weight AI is America's best bet, right as Chinese labs matched US benchmarks on capability, not just price. Read together, the letter looks less like confidence and more like a hedge against a coordination problem the signatories haven't actually solved.

2026-07-29

AIFeatured2 min read

AI Platform Strategy

Who controls AI infrastructure: four very different answers just arrived

From Apple's curated ecosystem to ElevenLabs' European independence, four distinct strategies are vying for control of AI infrastructure. Developers and enterprises face a fragmented field with no single winner.

2026-07-29

Labs & Research1 min read

Analysis

Benchmarks, not bargains: China's AI labs reach parity

Chinese AI labs Qwen, DeepSeek, MiniMax, and Kimi have matched or beaten US frontier models in reasoning, coding, and document parsing benchmarks. With aggressive pricing and open-weight strategies, they are no longer just cheap alternatives, they are real competitors.

2026-07-27

AIFeatured3 min read

Artificial Intelligence

41 companies, one message: open-weight AI is America's best bet

A rare coalition of 41 organizations, including OpenAI and Meta, makes the case that open-weight models are essential to American AI leadership. With MiniMax M

2026-07-27

Open SourceFeatured3 min read

Open source

Gemma 4 hit 300 million downloads. The math that broke the cloud model is why.

Google DeepMind's Gemma 4 reached 300 million downloads, a milestone signaling that the open-weight model's economic argument is finally beating the proprietary API pitch. Architectural efficiency and self-hosting costs are driving the shift.

2026-07-24

LLMs & ModelsFeatured4 min read

Google DeepMind

Gemma 4 just made every other open-weight model look 10x too big

Google DeepMind's Gemma 4 natively multimodal open-weight family introduces thinking mode, encoder-free architecture, and MoE options. The 2.3B model matches Gemma 3's 27B performance. The 31B model tops open-weight leaderboards.

2026-07-13

Labs & Research5 min read

Portfolio strategy autopsy

Mistral killed half its model family. What survived tells you everything.

A mid-2026 audit of Mistral AI's model portfolio reveals 36 active models across frontier, specialist, and legacy tiers, with 19 models slated for deprecation. The company's strategy emphasizes small specialist models for agents and coding, while retiring experimental variants like Magistrate and earlier Devstral versions.

2026-07-13

Google DeepMindFeatured4 min read

Google DeepMind

Google DeepMind's Gemma 4 turns 26 billion parameters into a reasoning machine that fits on one GPU

Google DeepMind's Gemma 4 technical report details a family of open-weight models with mixture-of-experts, 1M-token context windows, and multi-modal vision. The release signals a strategic play to bring frontier-level reasoning to developers without the cost of proprietary APIs.

2026-07-09