Changelog
What’s new
Public release notes, newest first.
v0.11.0
improvement- Server-side filtering added to the Google Ads and Meta Ads tools, so large accounts return the rows that matter instead of hitting row limits.
- DataLens Cloud is now allowlist-only, sourced live from the connection catalog — fewer stale or broken source options.
v0.10.0
feature- Added a Custom Server provider option, so teams running their own OpenAI-compatible endpoint can point DataLens at it directly.
- BigQuery multi-project support, with per-project safety checks before a query runs.
- Trial accounts default to an open-weights model, with existing trials backfilled automatically.
v0.9.0
improvement- Ads analytics now use pooled ratios for more stable percentage metrics on smaller accounts, plus an improved funnel visualization.
- Improved pacing, spend caps and error handling across the ad-platform integrations.
v0.8.0
feature- Gemini service-account authentication, including automatic model discovery for connected projects.
- BigQuery service account support as an alternative to OAuth for teams that prefer it.
- Provider logos added throughout the model picker for faster recognition.
v0.7.0
change- Simplified how DataLens maps your schema to query it — faster to extend and easier to reason about.
- Google Ads credentials moved fully server-side, with support for importing an existing google-ads.yaml.