Varun Mohan, a member of the team behind Antigravity, Google's coding-agent product, says the context windows of the three models in Antigravity (Argon, Carbon and Hydrogen) have doubled.

Mohan reports that compactions triggered per user turn fell by a factor of 3.5, but the post gives no window size before or after the doubling. When the window fills up, the system has to compress or drop part of the earlier content, and details get lost. The post gives no success rate showing whether fewer compactions make long tasks succeed more often, and doesn't explain how the 3.5 was measured. Google's Gemini 4 Argon model card says the model's own window limit is 1 million tokens of input and output combined; that is a model-level limit, not the setting inside Antigravity (Varun Mohan, 2026-11-16; Gemini 4 Argon model card, 2026-10). How we checked →
Source and checks
This item ran in the brief of Nov 16, 2026, alongside 10 other items. How we checked it, the quote checks and what was still missing that day are on that issue's How we checked page.
Cite this item
SecondSource (2026). "Varun Mohan, a member of the team behind Antigravity, Google's coding-agent product, says the context windows of the three models in Antigravity (Argon, Carbon and Hydrogen) have doubled.." https://secondsource.io/en/product/2026-11-16-1/. Accessed access date.