Model Releases
Ministral 3 8B: The Open-Source Vision Frontier
Mistral AI releases Ministral 3 8B, a powerful multimodal model with Apache 2.0 licensing that rivals frontier architectures.
5 min read
Latest news, tutorials, and insights about AI
Mistral AI releases Ministral 3 8B, a powerful multimodal model with Apache 2.0 licensing that rivals frontier architectures.
Mistral AI's newest 3B parameter model brings vision capabilities to local edge devices under Apache 2.0. Perfect for phones and drones.
Amazon unveils Nova 2 at re:Invent 2025, bringing MoE architecture and multimodal capabilities to enterprise via Bedrock.
Mistral AI unveils Large 3, a 41B active parameter MoE model with open weights, challenging global leaders in reasoning and efficiency.
Zhipu AI releases GLM-4.7, an open-weights coding model topping leaderboards with a cost-effective Flash variant.
MiniMax releases M2.1, a 230B MoE coding model offering SWE-bench scores of 74% at 92% lower cost than Western competitors.
Anthropic releases Claude Opus 4.5, a milestone reasoning model that outperforms Sonnet 4.5 by 4.3% with significantly lower token usage.
Allen AI releases OLMo 3, a fully open 32B parameter model with weights, data, and training code. Compare benchmarks, pricing, and architecture now.
Deep Cogito unveils Cogito v2.1, a groundbreaking 671B parameter MoE model designed for complex reasoning tasks. Explore benchmarks, pricing, and architecture details.