Latest news, tutorials, and insights about AI
IBM unveils Granite 4.0, a groundbreaking open-source model combining Mamba-2 and Transformer architectures under an Apache 2.0 license for maximum developer flexibility.
Anthropic's fastest model delivers near-frontier intelligence at unprecedented speeds with a 200K token context window.
DeepSeek AI releases V3.2 with 671B MoE architecture, rivaling proprietary models with open weights and enhanced benchmark performance.
Anthropic launches Claude Sonnet 4.5, claiming superior coding performance with a 1M token context window and advanced agentic capabilities.
Alibaba Cloud unleashes Qwen3-Next, an ultra-efficient 80B MoE model with only 3B active parameters, now available under Apache 2.0.
Moonshot AI launches Kimi K2, a massive 1T parameter model with 32B active weights, marking a historic open-source milestone with frontier-level performance.
Developers: xAI launches Grok 4 Fast with 98% cost reduction and real-time X search. See pricing, benchmarks, and API access details.
Mistral AI releases Magistral Medium 1.2, adding vision capabilities to its 45B parameter reasoning engine with a closed API model.
Mistral AI releases Magistral Small 1.2, a 24B multimodal reasoning model with Apache 2.0 license. Performance benchmarks and pricing details.