Chronological change log
A timeline of verified model launches, normalized into comparable change signals. Each entry links to the radar brief and the first-party source.
Developer capability
DeepSeek’s October refresh of its V4.1 Flash model for fast reasoning and coding workloads.
Capability expansion
OpenAI’s updated real-time voice model for responsive spoken interaction.
Capability expansion
An expressive GPT-Live-1.1 variant for engaging real-time conversation.
Capability expansion
xAI’s third-generation voice model for real-time spoken interaction.
Developer capability
Anthropic’s fastest small model for high-volume, cost-sensitive routing, summarization and subagent work.
Capability expansion
Mistral’s trillion-parameter, natively multimodal flagship preview for coding, agents and enterprise work.
Capability expansion
Cohere’s frontier embedding family with Pro and Fast variants sharing one multimodal, multilingual space.
Capability expansion
Near-Astra coding, computer-use and professional-work intelligence at one-fifth of Astra’s token price.
Capability expansion
A faster, lower-cost Sonnet upgrade for everyday work, coding and polished document creation.
Capability expansion
OpenAI’s balanced GPT-6 tier for capable coding, computer use and professional work.
Capability expansion
OpenAI’s fastest and most economical GPT-6 tier for focused, high-volume workloads.
Model launch
Anthropic’s latest Opus model for long-running agentic coding and knowledge work, with always-on adaptive thinking.
Model launch
SpaceXAI’s new frontier model for coding, agentic tasks and knowledge work, with stronger long-horizon execution and self-verification.
Capability expansion
Google’s real-time Gemini model for low-latency multimodal conversation.
Capability expansion
A Gemini 3.8 Live variant combining real-time interaction with extended reasoning.
Capability expansion
DeepSeek’s smaller V4.1 model with native visual understanding and a higher-throughput architecture.
Weights opened
An open-weight mixture-of-experts translation model covering more than 50 languages.
Capability expansion
OpenAI’s flagship GPT-6 model for demanding coding, computer-use, research and professional workloads.
Capability expansion
Google’s latest low-latency Gemini tier for high-volume multimodal and agentic workloads.
Capability expansion
A security-focused Gemini 3.8 Flash variant for authorized cyber defence and vulnerability workflows.
Capability expansion
Anthropic’s updated Fable tier for sustained frontier coding and demanding knowledge work.
Capability expansion
Anthropic’s updated Mythos tier for frontier reasoning, coding and knowledge work.
Weights opened
Z.ai’s efficient GLM-5.3 variant for strong coding and agentic performance with lower compute.
Capability expansion
DeepSeek’s experimental V4 Flash variant with visual understanding.
Weights opened
Z.ai’s open-weight flagship for complex coding, long-horizon agents and advanced cyber reasoning.
Weights opened
An open-weight model that composes, arranges, performs and produces complete songs.
Model launch
Google’s latest Flash model is generally available in the Gemini API with improved multimodal and coding performance.
Model launch
xAI’s latest Grok model release, announced by SpaceXAI on X as the successor to Grok 4.5.
Weights opened
A general-purpose multimodal model that understands text, images, video and audio and generates video with stereo sound.
Capability expansion
Google’s newest Flash model, tuned for token-efficient coding, knowledge work and multimodal tasks.
Developer capability
A fast frontier model built around coding, agentic execution and professional knowledge work.
Capability expansion
Meta’s multimodal reasoning model for agentic tasks, computer use, coding and rich media understanding.
Capability expansion
A three-tier family: Sol for frontier work, Terra for balance and Luna for fast, efficient tasks.
New interface
An embodied navigation model that follows language instructions using a single ordinary RGB camera.
Capability expansion
Anthropic’s fifth-generation Sonnet model for autonomous agents, coding, tool use and knowledge work.
Weights opened
An open flagship model designed to hold context and execute over long software-engineering horizons.
Capability expansion
Anthropic’s fifth-generation frontier model for sustained coding, science, vision and demanding knowledge work.
Weights opened
Cohere’s first agentic coding model, built to run locally with a small active compute footprint.
Weights opened
A natively multimodal frontier model for coding and agents, built on MiniMax Sparse Attention.
Developer capability
A new API generation offered as V4-Pro and V4-Flash, with both thinking and non-thinking modes.
AI Models Radar
