NVIDIA released Nemotron 3.5 Bolt E4B, an open-weight model optimized for fast local AI assistants NVIDIA
Moonshot AI released Kimi K3.1 Pro with a 1M-token context window and improved coding capabilities Moonshot AI
MiniMax released M3.1 Flash, an efficient AI model for everyday tasks MiniMax
Google released Gemini 4 Argon with a 1M-token context window for coding, complex reasoning, and autonomous agents Google
Anthropic released Claude Haiku 5.5, its fastest small model, with approximately 75% lower operating costs than Haiku 4.5 Anthropic
OpenAI released GPT-6.1 Sol with improved coding and agent capabilities at one-fifth of GPT-6 Astra's standard API token prices OpenAI
Anthropic released Claude Sonnet 5.5, running over 30% faster than Sonnet 5 and costing up to 30% less for most work Anthropic
Anthropic released Claude Opus 5.5 with stronger agentic coding capabilities and 40% lower typical costs than Opus 5 Anthropic
OpenAI released GPT-6 Sol and Luna, cutting API token prices by 50% compared with their GPT-5.6 promotional pricing OpenAI
Xiaomi released MiMo-V2.6 Pro and Flash with an Ultraspeed option Xiaomi
Anthropic introduced Claude Fable 5.1 for general use and Mythos 5.1 for restricted-access research in cybersecurity and life sciences Anthropic
DeepSeek released V4.1 Flash with native image understanding and a new architecture designed for faster inference DeepSeek
OpenAI introduced GPT-6 Astra for complex coding, research, and computer-use tasks, initially rolling it out to a limited set of organizations OpenAI
Google released Gemini 3.8 Flash for long-horizon software engineering, autonomous agents, and complex workflows Google
DeepSeek made V4 Pro generally available across its app, website, and API, with improved agent capabilities DeepSeek
Anthropic released Claude Opus 5 with improved long-running agent performance, coding, and professional work capabilities Anthropic
Google released Gemini 3.6 Flash for production use with improved token efficiency and code and agentic planning capabilities Google
Moonshot AI released Kimi K3 with a 1M-token context window, native vision, and long-horizon coding capabilities Moonshot AI
OpenAI released GPT-5.6 Sol, Terra, and Luna with improved coding and professional-work performance across different cost tiers OpenAI
Anthropic released Claude Sonnet 5 with improved planning, coding, and autonomous tool use at Sonnet-tier prices Anthropic