#open-source AI

All articles tagged with "open-source AI"

DeepSeek released preview versions of V4 with longer-context processing.
The company is offering Pro and Flash variants aimed at different workloads.
The launch reinforces China’s role in the global market for open-weight frontier AI.

DT Editorial Team·Apr 25, 2026·via technologyreview.com

India's Bhashini project unveiled a handheld AI device that runs models locally without cloud connectivity
Device supports 24+ Indian languages, addressing a gap in major commercial AI platforms
Open-source hardware and software stack could enable sovereign AI across developing nations

DT Editorial Team·Mar 15, 2026·4 min read·via restofworld.org

Alibaba released four Qwen 3.5 models including Flash, 35B-A3B, 122B-A10B, and 27B variants
The models use mixture-of-experts architecture to deliver high performance at lower compute cost
Alibaba claims performance competitive with GPT-5 mini and Claude Sonnet 4.5
All models are released as open weights, challenging the closed-source model paradigm

DT Editorial Team·Feb 26, 2026·4 min read·via the-decoder.com

Nous Research releases NousCoder-14B, achieving 42.8% on SWE-bench Verified versus Claude Code's 54% at a fraction of the size
The 14B parameter model runs on consumer GPUs with 24GB VRAM under Apache 2.0 license for unrestricted commercial use
The narrowing gap between open-source and proprietary coding models raises questions about the future of paid AI coding tools

DT Editorial Team·Feb 11, 2026·4 min read·via nousresearch.com

NousCoder-14B achieves 67.87% on LiveCodeBench v6, a 7-point improvement over its Qwen3-14B base model through reinforcement learning
The entire training pipeline, RL environment, and benchmark suite are published for full reproducibility by any researcher
Training took just four days on 48 Nvidia B200 GPUs using Nous Research's Atropos reinforcement learning framework

DT Editorial Team·Feb 11, 2026·3 min read·via venturebeat.com