# Spider Changelog - New Features and API Updates

What's new in Spider: recent features, API updates, and platform improvements.

Canonical: https://spider.cloud/changelog/

## What we've been shipping.

Recent product updates, API additions, and platform improvements.

## September 2026

Latest

1. Sep 10 New

   ### Provider router

   Put a router object in a scrape or crawl body to ask for an outside provider on your own keys, or set mode to off to keep a URL on Spider's own stack. Set funding to own to keep a request to routes your keys can pay for. Rolling out; write to support\@spider.cloud to check access. See the [provider router](/api/router/) page, or the [router docs](/docs/api/router/) for every field.

## March 2026

1. Mar 5 New

   ### Browser AI commands

   Send AI commands directly through Spider's Browser API WebSocket. Three modes: Act (click, type, navigate), Extract (pull structured data from the visible page), and Observe (describe what's on screen). Requires an [AI Studio](/ai/) subscription.

2. Mar 3 Update

   ### Dashboard logs redesign

   Redesigned the logs tab with a dark theme, click-to-expand detail panels, and an inline browser preview for live sessions.

## February 2026

1. Feb 25 New

   ### Scraper catalog

   1,000+ ready-made scraper examples across 32 categories. Each includes working code and a report button if a scraper breaks.

2. Feb 20 New

   ### Case studies and showcase

   Added [case studies](/case-studies/) with real production numbers and a [showcase](/showcase/) of verified integrations built with Spider.

3. Feb 15 New

   ### Proxy location pages

   199 country-specific proxy pages with unique profiles. Useful if you need to verify geo-targeted content or route crawls through specific regions.

4. Feb 10 AI

   ### Smarter AI extraction fallback

   AI extraction now uses a two-phase fallback: a fast model handles most pages, and a more capable model kicks in for complex layouts. Happens automatically per page.

5. Feb 7 New

   ### Data connectors

   Pipe crawl results straight into S3, GCS, Google Sheets, Azure Blob, or Supabase as pages are found. Set `data_connectors` on any crawl request. Works alongside webhooks.

6. Feb 4 AI

   ### Better extraction quality

   Raised the max extraction rounds so the models have more room to produce clean structured output on complex pages.

## January 2026

1. Jan 22 New

   ### Relevance gate

   When you use a prompt with AI extraction, pages that aren't relevant to your query get skipped before they're billed. Big savings on broad crawls with targeted prompts.

2. Jan 20 New

   ### AI usage tracking

   Every AI extraction logs token usage and deducts from your balance automatically. Per-request costs show in the dashboard.

3. Jan 15 New

   ### /ai/unblocker

   New endpoint that combines AI extraction with the Unblocker in one call. No separate config needed.

4. Jan 8 New

   ### extraction_schema parameter

   Pass a JSON schema via `extraction_schema` and the AI output conforms to it. Works with every extraction model.

5. Jan 5 Update

   ### AI pricing breakdown

   Added a per-model token pricing table to the [AI Studio](/ai/pricing/) pricing page.

## December 2025

1. Dec 20 New

   ### /ai/search

   Search the web and scrape results in a single call. AI extraction runs on each result page automatically.

2. Dec 15 Update

   ### Dashboard refresh

   Reworked the dashboard layout, fixed the cost breakdown display, improved mobile usability, and added a light mode for the playground editor.

[github.com/spider-rs ↗ ](https://github.com/spider-rs)· Track every commit on GitHub
