NeuralStackAI Release Tracker is presenting the race among major AI labs as a live timeline, not a series of isolated product launches. The site says it tracks 221 frontier models from 11 companies, beginning with ChatGPT’s debut on Nov. 30, 2022. It also says the monthly cadence of major AI releases has roughly quadrupled since 2023.
The tracker’s homepage shows a crowded pipeline of expected releases across OpenAI, Anthropic, Google, Meta, SpaceXAI, DeepSeek, Mistral, Moonshot AI, Cursor, Z.ai, and Qwen. Several are marked as overdue, including OpenAI, SpaceXAI, Mistral, and Cursor, while others are listed as expected in the coming days or weeks. That mix of overdue and upcoming entries underscores how quickly release schedules have become part of the story.
At the top of the tracker’s latest-release view, the most recent tracked frontier model is Muse Spark 1.2 by Meta, released on Aug. 5, 2026. The site also highlights recent activity around other labs, showing that the field is not dominated by a single cadence or company. Instead, major updates are arriving across multiple model families and vendors.
Beyond the timeline itself, AI Release Tracker says each model page includes release date, parameter count, context window, model type, and benchmark scores when published. Its analytics page adds cumulative release charts, monthly release frequency, year-over-year totals, benchmark frontier progression, and release cadence by company. The result is less a static catalog than an evolving record of how the AI industry is moving.
Why this matters
The pace of release is now a strategic signal in its own right. When a tracker shows major labs shipping on overlapping timelines, it suggests that product iteration, not just raw capability, is becoming a central part of competition. That shift makes the market easier to watch and harder to ignore.
It also changes how outsiders read the industry. Investors, developers, and customers can see not only who launched what, but how often each company is able to land new models or updates. In a field where model names and versions multiply quickly, that visibility can help distinguish sustained momentum from one-off announcements.
The tracker’s own framing points to a broader reality: frontier AI is moving so fast that release cadence is becoming part of the benchmark. With expected launches queued across several leading labs, the next few weeks will add more entries to a timeline that is already dense. The question is no longer whether the release cycle is accelerating, but which companies can keep up.