Articles
Paying ChatGPT users get a more focused Sol model with a reasoning effort slider. Free users gain unlimited text chats but lose access to OpenAI's strongest reasoning model.
MiniMax's H3 generates 2K video with native stereo audio in a single forward pass, and the open weights are already on Hugging Face.
ByteDance's Seed research team released SeedRealtime on August 5, a native audio-visual full-duplex large language model that fuses sound, vision, and text within a single unified architecture. Unlike cascaded systems that chain separate modules for speech recognition, vision, and text-to-speech, SeedRealtime runs perception, understanding, and response generation in parallel over continuous multimodal streams. The model has already been deployed in ByteDance's Douyin and Doubao consumer apps,
Tencent announced global availability of its Hy3 large language model on August 5, expanding access beyond China through three channels: the WorkBuddy AI workspace, the Miora creative studio, and the Tencent Cloud TokenHub model-as-a-service platform. The rollout follows Hy3's initial release on July 6 and comes with a free access period on WorkBuddy through August 31. Hy3 uses a hybrid fast-and-slow-thinking Mixture-of-Experts architecture with 295 billion total parameters and 21 billion activ
SaferAI's independent evaluation finds Zhipu AI's GLM-5.2 performs near saturation on offensive cybersecurity benchmarks while refusing none of the dangerous tasks it was given. Frontier developers like Anthropic and OpenAI consistently refused the same requests.
Mistral AI released Shieldstral on Monday, a 3-billion-parameter open-weights safety classifier that matches or outperforms guard models up to seven times its size. The model is available under Apache 2.0 and runs on a single 16 GB GPU. What sets Shieldstral apart from typical guardrail models is its approach to content moderation. Rather than baking a fixed taxonomy of harm categories into the model weights — which forces developers to retrain whenever their safety requirements change — Shield
Liquid AI released LFM2.5-2.6B on Monday, a 2.6-billion-parameter model designed to run capable AI agents entirely on consumer hardware. The model is available on Hugging Face with open weights and targets on-device deployment across laptops and phones. LFM2.5-2.6B achieves 220 tokens per second on an Apple M5 Max and 113 tokens per second on an AMD Ryzen CPU, fitting within 2.5 GB of memory. Liquid AI positions it as competitive with models roughly four times its size on tool use, instruction
Anthropic's most capable public model is generally available again after a three-week export-control pause. The interesting part isn't the model — it's the classifier stack now wrapped around it.
The latest tool-use APIs make agentic workflows production-ready. The remaining gotchas are subtle — and almost all about your prompt, not the model.
While the discourse focused on consumer chatbots and benchmark drama, the real story was hundreds of regulated workloads quietly migrating to Claude. Here's why.
