Proportional context compaction budgets, full-window sizing, custom model media support, and leak fixes

Tuesday, 22 September 2026

AI summary

Written by AI from the official notes. Check them for exact details.

Context compaction and media support have been improved, along with various bug fixes.

Improved context compaction spreads user-request budgets across all prompts.
Default PDF and audio support added for custom models.
Fixed Ctrl+G editor launches to work with real terminal stdout.
Voice dictation recordings now cap at 3 minutes 30 seconds to avoid failures.
Resolved memory leaks and resource issues in conversations.

Why it matters: Users of Gemini and Antigravity should care for enhanced performance and reliability in their AI interactions.

Official notes on antigravity.google

Full release notes11 changes

Improves context compaction by spreading user-request budgets proportionally across all prompts and sizing truncation ceilings from the model's full context window, enables default PDF and audio support for custom models, routes Ctrl+G editor launches to a real terminal stdout, caps voice dictation recordings to avoid deadline failures, and resolves resource and memory leaks across conversations and background pollers.

Improved context compaction to spread its user-request budget across all captured prompts, so a long initial instruction is no longer truncated to a small fixed slice when the other prompts in the conversation are short.
Improved context compaction to size summary and truncation budgets from the model's full context window instead of the compaction trigger threshold, so compacted summaries and background-task lists are no longer prematurely cut short.
Fixed PDF and audio tool outputs and attachments failing with unsupported mime type errors on custom models configured in settings.json; custom models now accept PDFs by default and honor the modelFeatures media flags for images, video, PDF, and audio.
Fixed pressing Ctrl+G to edit the prompt in a full-screen terminal editor such as vim hanging with Vim: Warning: Output is not to a terminal; external editors now run against a real terminal stdout.
Fixed voice dictation sessions longer than 4 minutes failing with a deadline error and erasing the drafted transcription; recordings now automatically stop and finalize at 3 minutes 30 seconds.
Fixed the startup banner displaying a Google Cloud Project ID for consumer (non-enterprise) sign-ins.
Fixed a stack-overflow crash when loading or compacting conversations containing background-task, subagent-management, messaging, or scheduling steps.
Fixed forked conversations inheriting step output data from steps after the fork point, which could collide with new steps produced in the fork.
Fixed a background poller and timer leaking for every conversation session, slowly accumulating memory over long-running sessions.
Fixed canceled conversation and workspace creation requests continuing to provision resources in the background; aborted requests now stop immediately.
Fixed quitting the CLI taking around 5 extra seconds before the process exited; shutdown now cancels open streaming connections immediately instead of waiting for a forced timeout.

Full release notes11 changes

Improved context compaction to spread its user-request budget across all captured prompts, so a long initial instruction is no longer truncated to a small fixed slice when the other prompts in the conversation are short.
Improved context compaction to size summary and truncation budgets from the model's full context window instead of the compaction trigger threshold, so compacted summaries and background-task lists are no longer prematurely cut short.
Fixed PDF and audio tool outputs and attachments failing with unsupported mime type errors on custom models configured in settings.json; custom models now accept PDFs by default and honor the modelFeatures media flags for images, video, PDF, and audio.
Fixed pressing Ctrl+G to edit the prompt in a full-screen terminal editor such as vim hanging with Vim: Warning: Output is not to a terminal; external editors now run against a real terminal stdout.
Fixed voice dictation sessions longer than 4 minutes failing with a deadline error and erasing the drafted transcription; recordings now automatically stop and finalize at 3 minutes 30 seconds.
Fixed the startup banner displaying a Google Cloud Project ID for consumer (non-enterprise) sign-ins.
Fixed a stack-overflow crash when loading or compacting conversations containing background-task, subagent-management, messaging, or scheduling steps.
Fixed forked conversations inheriting step output data from steps after the fork point, which could collide with new steps produced in the fork.
Fixed a background poller and timer leaking for every conversation session, slowly accumulating memory over long-running sessions.
Fixed canceled conversation and workspace creation requests continuing to provision resources in the background; aborted requests now stop immediately.
Fixed quitting the CLI taking around 5 extra seconds before the process exited; shutdown now cancels open streaming connections immediately instead of waiting for a forced timeout.