88 results for “LLM”
Alibaba's Qwen team released Qwen3.8-Flash-Next, a 125B-parameter multimodal MoE model with 6B activated parameters and 51B N-gram embeddings. The open-weight preview of Qwen4's architecture cuts training cost to one-ninth of its predecessor while leading coding and office-task benchmarks.
Bloomberg reporting cited by TechCrunch says Stripe has finalized an agreement to buy AI gateway OpenRouter for more than $7 billion. Neither company has confirmed the deal; Stripe declined to comment on speculation.
Z.ai released GLM-5.2 on June 13 with 1M context, MIT weights, and Coding Plan access before benchmarks. Developers can evaluate it in agents today while GLM-5.5 remains entirely unconfirmed.
Cursor has been acquired by SpaceX, completing an April partnership and gaining access to the largest GPU fleet for model training. Grok 4.6 launches as the first joint model, matching GPT-5.6 Sol on key benchmarks and available now in Cursor.
DeepSeek Flash, Qwen3.8-Flash, and GLM-5.3-Flash still price near $0.15/MTok input — while Opus and Sol sit at flagship multiples. Brocker’s baseline table is refreshed from vendor pages as of 15 September 2026.
Z.ai published a blog post describing custom inference infrastructure for its GLM models, framed as a step toward recursive self-improvement. No performance data or hardware details were disclosed.
DeepSeek’s official page confirms V4.1-Flash is live as deepseek-flash with a 552B MoE encoder–decoder stack. Pro requests remap to Flash rates from 04:00 UTC on September 14 until V4.1-Pro.
DeepSeek V4 Flash is positioned as a fast, lower-cost open-weight AI model for developers who need long-context reasoning without paying frontier-model. The model is described as a Mixture-of-Experts system with 284 billion total.
A JPMorgan note relayed by Reuters and CGTN forecasts a Zhipu AI model exceeding 1 trillion parameters for August 2026. Zhipu has not confirmed the model, its specs, or the date — leaving developers to separate analyst signals from vendor commitments.
DeepSeek's next flagship model, DeepSeek V4, appears to be on track for an official launch on August 3, 2026, according to multiple converging signals. The competitive landscape DeepSeek V4 enters is markedly different from the one.
CISA added seven actively exploited vulnerabilities to the KEV catalog, including flaws in LiteLLM, Kestra, SonicWall SMA 1000, JFrog Artifactory, and Sangoma Switchvox. Microsoft and Wiz report that AI infrastructure gateways are now active targets for reverse shells, crypto miners, and API key theft.
PyTorch Foundation, Hugging Face, IBM, and Mitsubishi Electric will host PyTorch Day Japan 2026 in Tokyo on December 10.