Nov
2022
ChatGPT first released
OpenAI launches ChatGPT as a research preview. The moment LLMs crossed from research demo to mainstream utility, becoming the fastest-adopted consumer product in history.
New primitives are coming
Subspacelabs
Projects
Agent Test Range.
Subspacelabs is exploring test ranges for autonomous agents: simulated deployments, tool-call sandboxes, local inference paths, and audit trails for systems that need to act before they are trusted at scale.
Timeline
Feb
2025
Claude Code released
Anthropic previews Claude Code for agentic coding workflows. First credible move from "AI that suggests code" to "AI that does the work end-to-end."
Nov
2025
OpenClaw released
OpenClaw first appears publicly as an open-source local agent project. Showed the value at the intersection of self-improvement, broad permissions, and easy access in a single agent.
Feb
2026
Opus 4.6
Anthropic ships Claude Opus 4.6. Introduced "agent teams" and a 1M-token context, pushing coding agents from single runs to sustained, multi-agent work.
Feb
2026
GPT-5.3-Codex
OpenAI introduces GPT-5.3-Codex. Set a new SWE-Bench Pro high and was the first OpenAI release flagged as "High" capability in cybersecurity.
Feb
2026
Qwen3.5
Alibaba officially releases Qwen3.5. Made frontier-tier intelligence something you can download, not just rent — reshaping the open-vs-closed balance.
Mar
2026
TurboQuant
Google Research introduces TurboQuant. Compressed KV caches 6× with near-zero accuracy loss and no retraining, easing the memory bottleneck behind long-context inference.
Apr
2026
Gemma 4
Google launches Gemma 4. A 31B open model that matches 400B-class rivals, with multimodal reasoning running fully offline on phones and Jetson-class hardware.
May
2026
Gemini Omni
Google introduces Gemini Omni Flash, merging Gemini's reasoning with video generation and natural-language editing. Multimodal models move from answering prompts to shaping media directly.
Jun
2026
Claude Fable 5 / Mythos 5
Anthropic launches Mythos-class models with frontier autonomy across coding, science, vision, and cyber. Days later, access is suspended, turning model release into an infrastructure and governance event.
Jun
2026
DiffusionGemma
Google releases DiffusionGemma, an open text-diffusion model built for fast local and interactive generation. The bottleneck shifts from raw intelligence to latency, hardware fit, and new inference patterns.