Five new endpoints give you access to the Global Macro Sentiment Indices (GMSI) dataset — hourly country-level sentiment aggregates across 70 macro topics, built from a dedicated country-matched article pipeline with separate directional and semantic sentiment models.
You can now retrieve paginated historical smoothed trading signals for any ticker using GET /v1/signals/historical/{ticker}. The endpoint applies the same hourly resampling and smoothing algorithm as the live GET /signals endpoint, so the historical series is consistent with the live signal you already receive.
The signals endpoint now supports a new v3 copilot analyst forecast model, available both with (v3) and without (v3_raw) additional risk management logic. The latest alias has been updated to resolve to v3. The default value of model_name remains v2 for backwards compatibility.
18 endpoints that were deprecated in v1.5.0 and v1.6.0 and scheduled for removal on 2026-05-01 have been permanently removed. If you are still calling any of these endpoints, migrate to the replacements below.
The Signals API now includes an ensemble endpoint for retrieving the latest ensemble trading signals for a single ticker. This endpoint is currently available for 'BTC_CRY' only.
A new family of dataset endpoints is now available, giving API clients direct access to the Permutable data catalogue — including dataset listings, metadata, versioning, and download links.
The macro regional sentiment endpoints now support an interval parameter (HOUR or DAY) to control the granularity of the time-series response. The legacy historical data endpoints have been removed.
The signals endpoint now queries a 3-day lookback window instead of 14 days when computing the latest smoothed signal, significantly improving response times for requests with many tickers.
The signals endpoint now supports eight additional model types — including a latest alias, non risk managed forecasts for both v1 and v2, technical forecasts, as well as individual & aggregated sentiment variants — and exposes a configurable smoothing parameter, giving you finer control over signal stability versus responsiveness.
Topic preset and individual topic filtering is now available directly on all headline feed and asset index endpoints, removing the need to filter topic data client-side.
