> For the complete documentation index, see [llms.txt](https://changelog.graphlit.dev/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://changelog.graphlit.dev/september-2025/readme-13.md).

# September 30: Support for ingestion performance, search stability, and token accounting accuracy

### New Features

* :bulb: This release includes broad **performance optimizations** across ingestion, search, and summarization pipelines. Developers should see faster indexing for large document batches and improved throughput when embedding concurrent content streams.
* We’ve also refined **token accounting and reporting** within the billing subsystem, ensuring real-time accuracy for both input and output token usage across all connected model providers.

### Bugs Fixed

* GPLA-4590: Resolved high-concurrency queue deadlocks affecting ingestion of large feed batches
* GPLA-4594: Improved summarization retry logic for timeouts on long transcripts
* GPLA-4601: Corrected mis-reported embedding-cost metrics for Gemini models