Tag
112 articles
This article explains Google's new Gemini 3.5 Transcribe speech-to-text model, detailing its dual-endpoint architecture, technical mechanisms, and implications for developers building voice agents and transcription systems.
Z.ai confirms it is behind Ox Alpha, the mysterious open AI model topping benchmarks and leaderboards, with its weights set to be released soon.
Alibaba's Qwen team introduces Qwen3.8-Flash-Next, a cost-efficient model that uses only 6% of its parameters per token, outperforming competitors like Claude Opus 4.6 and DeepSeek-V4-Flash.
Alibaba's Qwen team introduces Qwen3.8-Flash-Next, a 125B multimodal MoE model with only 6B active parameters per token, showcasing significant training efficiency and architectural innovation.
Generalist AI has released GEN-1.5, a robot foundation model that learns new tasks from a single 3–12 second demonstration, without requiring fine-tuning or gradient updates.
Anthropic has revealed it is using an unpublished, highly capable AI model internally, codenamed Model 2, which surpasses any publicly available version of Claude.
GLM-5.3, Zhipu AI's latest open-source model, has topped performance rankings and undercut rivals on price, though its release has been delayed.
China's Z.ai has released a powerful new AI model that could enhance cybersecurity but also poses risks if misused by hackers.
Apple has reportedly developed a custom AI model for the Chinese market in collaboration with Alibaba, marking a rare cross-border tech partnership amid U.S.-China tensions.
Explore the advanced AI concepts behind Needle 2, a 45-million-parameter model that runs efficiently on minimal hardware. Learn how tool calling, quantization, and edge deployment are revolutionizing AI accessibility.
Learn how Z.ai improved their AI model GLM-5.3 without rebuilding it from scratch, by giving it more targeted training. Understand the concept of scaled post-training and its impact on AI performance.
Writer introduces a new AI model built on Z.ai's GLM-5.2 that offers deployment-ready capabilities at significantly reduced token costs.
