
Learn about Merge’s latest integrations and features
Each adds a write endpoint for core financial records.
Read more in the Accounting docs.
Improvements
Agents can now route requests to models built for audio or video input through the same Gateway endpoint they already use for text and image models. Read more in the Gateway docs.
Each project can now set its own Prompt Injection Protection thresholds and DLP rules, overriding the organization default where it needs to. A project can only tighten the organization's policy, never loosen it, so the org-wide setting stays the floor for every project under it. Read more in the Gateway security docs.
Improvements
Outlook message bodies now return as plaintext by default and drop out of list results entirely: a single message shrinks from around 57kb to 14kb, and a ten-message list from 394kb to 16kb. Calendar tools still offer an HTML opt-in for when an agent specifically needs rich formatting. Read more in the Outlook docs.
File lookups and search resolve Google Drive shortcuts to their target file, so agents read the real document without a second call. Breaking change: if your code inspected shortcut metadata directly, update it before upgrading. Read more in the Google Drive docs.
Billing now keeps a running record of credit consumption over time for enterprise admins, not just a current snapshot. Read more in the Agent Handler docs.
Improvements
A scheduled promotion or compensation change on SAP SuccessFactors now shows up in GET /employments before it takes effect, sitting alongside the employee's current record. Pull both the current and upcoming employment data in one call instead of tracking effective dates separately. Read more in the employments docs.
Improvements
You can now route to Kimi-K3 through US-based inference providers, with Makora and Modal added as supported vendors.
Read more in the Gateway docs.
Improvements
Merge now runs generic MCP connectors on the same authentication, access control, and governance infrastructure as its own, so enabling one takes the same steps as enabling a Merge-built connector, with no MCP server to stand up.
Read more in the Agent Handler docs.
You can now pull identity, spend management, spreadsheet, BI dashboard, analytics, feature flag, and meeting-note data through Agent Handler.
Read more in the Agent Handler docs.
Calendar reads target the primary calendar by default, so syncs stay fast on accounts with many calendars. To include a secondary calendar, name it explicitly in the request. Read more in the Google Calendar docs.
If you built on the FactSet connector, move those integrations to an alternative data connector. It is no longer available in the catalog.
Read more in the Agent Handler docs.
You can now see every guardrail rule and test it in one place, instead of checking the rules and their behavior separately.
Read more on the Agent Handler page.
You can add a card and start billing directly from the dashboard, with no sales or support handoff.
Read more on the Agent Handler page.
Improvements
You can now manage scopes directly on the page, one change at a time or many at once.
Read more in the Unified API docs.
You can now pull Hive tickets, comments, and attachments through GET /tickets, GET /comments, and GET /attachments.
Read more in the Ticketing docs.
Improvements
You can now send one prompt to a panel of models in parallel, then have a judge model synthesize their answers into a single response.
Read more in the Merge Fusion docs.
You can now set indirect prompt injection protection per axis, picking a mode (off, alert, or block) and a sensitivity threshold for each one.
Test the configuration against injection patterns with a built-in detection tester in Security settings before it goes live. Read more in the prompt injection protection docs.
You can now set a service_tier request parameter to toggle between flex and priority processing, trading cost for latency or the reverse.
Per-tier pricing and billing now show up directly on /v1/models, so the tradeoff is visible before you pick a tier. Read more in the service tiers docs.
The Google Ads connector adds 22 tools across the full account structure.
Tools cover customers, campaigns, ad groups, ads, budgets, and keywords, plus a dedicated GAQL tool that runs Google Ads Query Language queries directly.
Read more in the Agent Handler connector docs for the full tool list.
Agents can now create, read, update, delete, and search HubSpot tickets through the same connection as HubSpot CRM records. Ticketing runs as an optional scope alongside the existing CRM tools.
Adding a custom MCP server to the catalog takes one page instead of switching between screens: the Configure wizard imports the server and sets up its credentials in the same flow.
If the server doesn't require a client ID upfront, Agent Handler registers a client automatically, so there's no manual OAuth app to create first. Read more in the custom MCP servers docs.
You can now hide auth, configuration, and Selective Sync screens in Link for a single integration without changing behavior for any other integration.
Selective Sync keeps running in the background even when its screen is hidden.
Read more in the Merge Link docs.
GET /sync-status now returns two new fields on every category's SyncStatus object: sync_status_reason (RATE_LIMITED or WAITING_ON_OTHER_MODELS) and data_fresh_as_of.
Both ship in Python SDK 4.1.0, so you can see why a sync paused and how current its data is.
SharePoint and OneDrive syncs resume from their last checkpoint.
Read more in the time off type reference for the full list of time off types.