Welcome to Volume 1 of our 2026 product release notes! We've been working hard to bring you performance improvements that add value to your work. Read on to learn about the product enhancements and bug fixes released in January 2026.
Benchling AI
Benchling's artificial intelligence agents and features.
Added manual format button to SQL Writer [2026.1.AI.1]
SQL Writer no longer automatically formats suggested SQL queries, which previously created unnecessarily large diffs when making small query modifications. Users can now format SQL on demand using a new Format SQL button, preserving the original query structure until explicitly requested.
Updated back button styling in AlphaFold viewer [2026.1.AI.2]
Updated the back button in the AlphaFold structure prediction viewer to display 'Back' instead of 'Prediction List' and changed its styling from a button to a text link
Enhancements to Ask mode's warehouse querying [2026.1.AI.3]
Ask mode will now retry warehouse queries if its first attempt fails making it more likely to successfully answer questions it previously may have suggested users use Deep Research to answer.
Updated Deep Research's understanding of Benchling Projects and Folders [2026.1.AI.4]
In the warehouse, projects are represented both as top level folders and permissions sources. Deep Research and Ask now have a better understanding in deciding which to use for what to avoid retries.
Enhancements to Deep Research Citations [2026.1.AI.5]
Enhancements to help Deep Research better handle citations on longer outputs.
Deep Research System Table Understanding [2026.1.AI.6]
Deep Research now has a better understanding of the Benchling-generated, system warehouse tables and how to connect them for common use cases. This should reduce retries and enhance performance.
Enhancements to help Deep Research handle reading very large notebook entries [2026.1.AI.7]
Enhancements to help Deep Research handle reading larger notebook entries more efficiently
Better handling of large files when using Data Entry Agent within Compose [2026.1.AI.8]
Benchling adjusted Data Entry Agent to better handle larger files and to ask for clarifications when called via Compose to help fill data in structured tables.
Increased limits for Deep Research Citation Agent [2026.1.AI.9]
Increased the limits available to the Deep Research Citation feature to decrease the likelihood that a longer report would complete but adding citations would fail.
Data Entry Agent, if enabled, is available to the Compose Agent to use as a tool [2026.1.AI.10]
Data Entry Agent within the Compose agent allows Compose to create a Notebook Entry or template based on a file and then use Data Entry Agent to fill out the data in any structured tables in the entry or template.
Compose Agent can fill data from files it is instructed to find [2026.1.AI.11]
In addition to filling data into tables from the files provided to the agent, Compose can now follow instructions from its prompt to find data and then use those data to fill tables in the draft entry.
Compose and Deep Research show progress updates [2026.1.AI.12]
Compose and Deep Research now display real-time progress indicators showing which tools are running, such as filling tables or searching files, and any "thoughts" about the approach the agent will take. Previously, users would see an unchanging screen during long operations. Progress sections remain visible after completion for reference.
Checkboxes in Compose Agent [2026.1.AI.13]
The compose agent can now insert and edit checkbox list items.
Compose agent can insert container creation inventory tables [2026.1.AI.14]
The compose agent can now update documents with the "Create and fill new containers" inventory table
Subtemplate Insertion in Compose [2026.1.AI.15]
The Compose agent can now insert subtemplates into draft documents. Users can provide a subtemplate API ID to copy reusable content blocks (such as standard protocols or experimental setups) directly into their documents.
Inventory table support in Compose [2026.1.AI.16]
Compose can insert inventory tables for updating existing containers
Users can continue from failed chats [2026.1.AI.17]
In some situations a chat could encounter an error that prevents sending additional messages in the same thread. This update allows users to send a new message continuing from before the error and removing the errored message.
Remove repeated tool calls on follow up messages [2026.1.AI.18]
Previously, Benchling would show the entire tool call and thought history from the entire chat thread on each follow up message. Now Benchling only shows the new tool calls for each response between the user question and the agent response message.
AI Agent Chat Threads now have user downloadable audit logs [2026.1.AI.19]
AI Agent Chat Threads now have user downloadable audit logs available in the gear icon
Updated Agent naming in tenant admin interface [2026.1.AI.20]
Updated AI feature names in the admin interface for consistency. Compose is now Compose Agent, Deep Research is now Deep Research Agent, and related settings use Agent terminology throughout. This improves clarity when configuring AI capabilities.
Changed the rate limit for AI feature use for each user to per hour instead of per day [2026.1.AI.21]
Benchling implemented rate limits for each user across all AI features to prevent abuse. These limits were adjusted to reset hourly instead of daily.
Increased per tenant hourly rate limits for AI features [2026.1.AI.22]
Benchling implemented rate limits that apply across each customer tenant across all AI features that limit total AI feature use per hour to prevent abuse. These rate limits were increased across all tenants
Updated Notebook Check Models [2026.1.AI.23]
AI models used by the Notebook Check feature were updated
Updated SQL Writer models [2026.1.AI.24]
AI models used by the SQL Writer feature were updated
Updated Compose and Data Entry Agent Models [2026.1.AI.25]
AI models used by the Compose Agent and Data Entry Agent were updated
[Beta] Deep Research supports MCP Connectors [2026.1.AI.26]
This feature allows tenant admins to add MCP connectors to other common software systems to allow Deep Research to retrieve information from those systems in addition to Benchling data to generate its response.
[Beta] Benchling Remote MCP open beta [2026.1.DP.27]
A Remote MCP is now in open beta. This MCP can easily be installed by admins to chat clients like Claude. The MCP includes "Ask" and "Deep Research" as tools, allowing users to interact with those agents from any chat client that supports MCPs.
Notebook
Benchling Notebook helps users log their experiments, create protocols, and capture results against their samples.
Improved text insertion in tables within notebook entries [2026.1.N.1]
Tables inserted via the insert menu now automatically include a new line below, making it easier to add text between tables and other elements like section headers.
Updated display for review process name in the review summary panel [2026.1.N.2]
The review summary panel now displays the name of the review process used, helping reviewers identify which process applies in projects with multiple review processes
.
New table navigation keyboard shortcuts and modal [2026.1.N.3]
Added new shortcuts for table navigation (i.e. adding blank rows, navigating to the start/end of columns and rows, selecting rows and columns), as well as a modal to show Notebook and Table shortcuts.
Cross-Step Data Referencing in Worksheets [2026.1.N.4]
Cross-Step Data Referencing enables formula references between tables in different steps within the same worksheet.
Users can create formulas like =Step1!Measurements!A5 to connect data across multi-step workflows, reducing data entry and maintaining integrity when source data changes. Note that multi-user editing is disabled when this feature is enabled.
Email Benchling Support to enable
Entry chip popover shows display ID and creation date [2026.1.N.5]
The Entry chip popover now shows additional details including the entry's display ID and when it was created.
Template collections in the workspace [2026.1.N.6]
Template collections are now accessible directly from the main workspace, making it possible to edit templates in the context of other work. Templates open in tabs alongside other workspace items like studies and entries, with improved search, familiar filters, clearer visual distinction from instances, and expanded favoriting to keep important templates and collections easy to find.
Email Benchling Support to enable
Relevant Items now include Worksheets [2026.1.N.7]
When an entity is referenced in a worksheet, that worksheet will now appear in the entity's Relevant Items tab — just like notebook entries do today.
Molecular Biology
Benchling's Molecular Biology is a cutting-edge in silico design tool that supports the latest scientific workflows.
Amino acid search supports exact match filtering [2026.1.MB.1]
Users can now search for protein sequences using exact amino acid string matches in addition to the existing partial match option. This enables more precise filtering when searching for proteins with specific amino acid sequences.
Improved NCBI sequence import reliability [2026.1.MB.2]
Updated NCBI API integration to use new API key; this reduces timeout errors and improving reliability when importing sequences from NCBI databases.
Resolved BLAST Timeout Issues [2026.1.AC.2]
Addressed a limitation where extensive BLAST queries would exceed session limits. A new pipeline now ensures successful completion for long sequences even over 20k.
[Beta] Configurable rules for biointelligent link "autofill" (when overlapping/new parts is enabled only) [2026.1.MB.3]
For customers who have new ("overlapping") parts enabled, we released configurable autofill and validation rules for part, transcription, and translation links. This feature gives customers more control over which links are filled and reduces noise and time spent on complex, scaled sequence workflows.
Email Benchling Support to enable
In Vivo
Manage in vivo studies with modern software designed for data capture, collaboration, and insights.
[Beta] Task Specs API and webhooks (Beta) [2026.1.IV.1]
Introduced a new set of beta API endpoints for managing task specs for tasks of type "Other". These endpoints allow programmatic creation, retrieval, updating, and deletion of task specs, along with management of assignees and targets, providing the same functionality available in the UI.
This initial release supports the "other" task type only; additional task types (measurement, sample, observation, dosing) are planned for future releases. Note: This is a beta API intended for testing purposes. Endpoints and response structures may change in future releases.
New Endpoints:
- Task Spec Management
- Create a task spec — POST /api/public/v1-beta/studies/{studyId}/task-specs
- List all task specs (filterable by type) — GET /api/public/v1-beta/studies/{studyId}/task-specs
- Retrieve a task spec — GET /api/public/v1-beta/studies/{studyId}/task-specs/{taskSpecId}
- Update a task spec — PATCH /api/public/v1-beta/studies/{studyId}/task-specs/{taskSpecId}
- Delete a task spec — DELETE /api/public/v1-beta/studies/{studyId}/task-specs/{taskSpecId}
- Assignee Management
- List assignees — GET /api/public/v1-beta/studies/{studyId}/task-specs/{taskSpecId}/assignees
- Add assignees — POST /api/public/v1-beta/studies/{studyId}/task-specs/{taskSpecId}/assignees/add
- Remove assignees — POST /api/public/v1-beta/studies/{studyId}/task-specs/{taskSpecId}/assignees/remove
- Target Management
- List targets (group targets only) — GET /api/public/v1-beta/studies/{studyId}/task-specs/{taskSpecId}/targets
- Set target groups — POST /api/public/v1-beta/studies/{studyId}/task-specs/{taskSpecId}/targets/set
- Webhooks (Contact Support to enable)
- Task spec created — study.task_spec.created
- Task spec updated — study.task_spec.updated
- Task spec deleted — study.task_spec.deleted
Registry and Inventory
The Registry and Inventory combine to form an extremely powerful solution for life sciences R&D samples.
Added concentration layer to plate maps [2026.1.RI.1]
This release enables users to assign concentration values directly on the plate by setting them manually or by designing a serial dilution and to visualize concentration on plate maps
Workflows
Workflows are a sequence of processes that are performed to answer a research question or synthesize new material.
Task studies pre-populate entry creation modal [2026.1.W.1]
When executing workflow tasks into a new entry, the study selection field now automatically prepopulates with studies associated with those tasks. Users can accept the prepopulated studies or change them before creating the entry, streamlining the workflow execution process.
Insights
Insights enables users to query, visualize, and share high quality, structured data that resides in Benchling.
Enable rendering of empty charts with no input data within Analysis [2026.1.IN.1]
Previous behavior was to produce an error of a failed chart. New behavior will be to create an empty rendered chart
Add minimum/maximum range configuration to X and Y axes of Line charts [2026.1.IN.2]
Added minimum/maximum range configuration options to both settings of X and Y axes of a Line chart created in Analysis. Users can set range boundaries for plotting of data with these charts
Look up schematized data in analyses and analysis templates [2026.1.IN.3]
Analyses and analysis templates can look up schematized data. To add a look up step to an analysis, click *Add table > From look up*. The schematized data will refresh whenever the analysis template is run. Entity schemas are currently available for all tenants; additional schema types will be supported in the future.
Enable drop-down selection of table-based data flagging reason [2026.1.IN.4]
Enable user configuration of pre-defined Table-based data flagging/unflagging reasons via configuration flags. Upon flagging/unflagging a new column "Reason" will be created within the table along with original table-based flagging behavior of free-text comments.
Rename flagging display column names [2026.1.IN.5]
Renamed data flagging display column names from "Comment" to "Flag Comment" and "Reason" to "Flag Reason" in analysis tables. This change makes column naming clearer when working with flagged data. Previously flagged rows remain editable with the updated column names
Custom Warehouse Views [2026.1.IN.6]
With custom views, users are able to define re-usable SQL queries that can be used via Insights or Direct Warehouse Access. These views are stored in the warehouse and they surface "fresh" data (i.e. the most up-to-date data that exists in the warehouse).
This enables users to make query writing much simpler--instead of having to copy and paste the same complicated query many times, a user can simply reference a custom view instead.
[Beta] Insights Labs: New visual flagging feature to manually flag and exclude points in regressions plots [2026.1.IN.7]
Key capabilities:
- Lasso & flag: Click and drag on any regression plot to select point(s) and flag, then add comments to document why they're interesting
- Exclude from fits: Toggle flagged points in/out of regression calculations to see how they impact your model
- Track flagged data: Flagged points appear with an "X" marker on charts and get automatic
- Flag/Flag Comment columns in your Source Table
- Compatibility with Table-Based Flagging: Whether you are flagging from table-based view or visually, points within the chart will be styled consistently
To Enable: Turn on Insights Labs via the Tenant Admin Console -> Settings settings tag ->Insights Lab. Advanced Analysis or BioP license is required.
Schemas
Schemas are the structural foundation for how your information is categorized in Benchling. They are categories created specifically for your tenant that can represent materials, data, storage methods, and tasks.
The field_definition warehouse table now includes target fieldset information [2026.1.SCH.1]
The field_definition warehouse table now includes a target_fieldset_id column for fieldset linked fields. This enables warehouse queries to directly identify which fieldset a linked field points to without joining additional tables.
Application Core
To create a seamless Benchling experience, our platform must be consistent, compliant and predictable. The Application Core provides the central capabilities to create a smoother user experience.
Updated default tab for Organizations [2026.1.AC.1]
Updated the default tab for the Organization page to Members list (used to be Projects)
[Beta] New config migration import UI [2026.1.AC.2]
Updated UI for config migration import to improve feature usability, efficiency, and support for more complicated features being imported. Instead of a stepped experience requiring multiple validation steps, all object types can now be matched and validated in a single step. When enabled, the feature provides an option for users to switch between the original and new UI as desired.
Email Benchling Support to enable
[Beta] New access policy settings for results [2026.1.AC.3]
There are new access policies to control users' ability to create and edit results, with more granularity.
Results > Create allows users to create results, but does not allow editing results once reviewed.
Results > Edit allows users to edit results, including reviewed results, through Benchling's V3 API. To prevent users from editing results through Benchling's API, set the policy statements under Results > Edit to Not granted.
Here is a detailed breakdown of the new access policies:
- Results > Create > Create results in Notebook entries
- Allows users to create, archive and update results in in-progress entries.
- Will be granted to users with Notebook entries > Edit other data access, which previously governed the same capabilities.
- Results > Create > Create results in worksheets
- Allows users to create, archive and update results in in-progress worksheets.
- Will be granted to users with Worksheets > Edit access, which previously governed the same capabilities.
- Results > Create > Create other results
- Allows users to create and archive results that are not in entries or worksheets.
- Will be granted to users with Projects and folders > Edit folder properties access, which previously governed the same capabilities.
- Results > Edit > Edit results that were created in Notebook entries, including reviewed entries
- Allows users to edit results that were created in Notebook entries, through Benchling's API. To edit results through Benchling's API, users additionally need to be granted access to the developer platform.
- Will be granted to users with an admin access policy. This includes the default admin access policy and access policies duplicated from the default admin policy.
- Results > Edit > Edit results that were created in worksheets, including reviewed worksheets
- Allows users to edit results that were created in worksheets, through Benchling's API. To edit results through Benchling's API, users additionally need to be granted access to the developer platform.
- Will be granted to users with Worksheets > Admin access.
- Results > Edit > Edit other results
- Allows users to edit results that are not in entries or worksheets, through Benchling's API. To edit results through Benchling's API, users additionally need to be granted access to the developer platform.
- Will be granted to users with Projects and folders > Edit project settings access.
Email Benchling Support to enable
Developer Platform
Our developer platform is fundamental to centralizing and standardizing all your R&D data. Benchling's APIs and Events are built to match the flexibility and speed of modern life science R&D.
[Beta] Review process information available in v2beta API [2026.1.DP.1]
Review process details, including stages, reviewers, and completion status, are now available through the v2beta API for entries and worksheets. This enables users to build integrations and automations that utilize review workflow information without using alpha endpoints.
[Beta] Bulk entries creation API [2026.1.DP.2]
The bulk entries creation API endpoint has been promoted from alpha to beta stability. Customers can now use this endpoint in production to create multiple notebook entries in a single API call.
[Beta] New V3 API to create and update results [2026.1.DP.3]
Now, results can be created, archived and edited with the V3 API. Single-update APIs are available in v3-beta and bulk update APIs are available in v3-alpha.
These types of results cannot be updated with the API:
- Results in legacy tables (i.e. results in tables without a "Submit" button)
- Results in legacy runs (i.e. results with a run link field)
PipeBio
Benchling's cloud-based bioinformatics software for antibody/peptide screening & drug development. Conduct BCR, TCR, VHH, scFv, or peptide sequence analysis and integrate your workflows with Benchling's core platform.
Support for IP restriction [2026.1.PB.1]
PipeBio now enables customers to define allowed IP addresses. Access from other IP addresses is blocked.
This feature is not on by default, it is enabled by admins.
Bug Fixes
Benchling AI
- Fixed a bug where Notebook citations could show up as N/A
- Fixed a bug case where switching from Ask to Deep Research could error
Notebook
- Datetime Auto-fill Intermittently Fails in Structured Tables
- Retract action unblocked for entries with unsubmitted tables
- Entity Link Search Fixed for First Keystroke in Registration Tables
- Fixed changed value indicators on study columns
- Auditor review will not block count failed computed cells
- Inventory table submission fixed
- Approved worksheet status message corrected
- Entry scroll position preserved when submitting tables
- Fixed Formulas Re-shift When Regenerating Process Intermediates
- LOOKUPFIELD configuration error fixed in legacy results tables
- Fixed column locking in plain tables
- Notebook table column menu options fixed
- Fixed copy/paste dropdown issue
- Fixed single input columns in CSV upload through entry creation API
- Worksheet review snapshots include audit logs correctly
- Image editor preserves rotations and allows resizing
- E-signature authentication handles blocked popups gracefully
- Worksheet exports now properly evaluate cross step formula references
- Entry version cloning now tracks parent lineage
Molecular Biology
- Fixed inconsistent amino acid count
- Registration table edits work without MolBio capability enabled
- Fix to DNA sequence entity link validation
Small Molecule
- Fixed oligo conjugate registration table bug
- Fixed custom notation error bug
In Vivo
- Corrected Restrictions for Number of Sample Plan Metadata Columns
- Study Code Sequence Reset Now Functions Correctly at Year Boundaries
- Observations on animal profile cannot be deleted
- Audit Log Now Shows Accurate Reason for Sample Cancellation
- Cages linked to deleted studies can be deleted from colony
- Data table exports include all column values
Registry and Inventory
- Fixed a bug for plate maps that didn't support adding from fieldset
- Fixed labeling strategies bug blocking config migration
- Fixed bug impacting plate icon in tables
Workflows
- Fix error on sending binary files to File Watcher connection
- Fixed legacy request field validation
Insights
- Allow deletion of Analysis views that are created from a Notebook table
- Fix issue with Heatmap gradient legends not displaying
- Fixed not being able to edit transforms of tables created from a notebook
Benchling Connect
- Improved File preview performance
- Improved destination folder selection
- Fixed Hidden Processor Outputs Affect Performance in Lab Automation Runs
- Improved destination folder selection
Schemas
- Template collection search no longer flashes loading indicator
Application Core
- Benchling Sync file check-in fixed
- Fixed Export Error for Location History
- Fixed issue in Permissions Report where 3 actions were not being surfaced properly (Request/Legacy Request creation, Location creation, and Batch creation)
- Fixed issue with config migration exports timing out on tenants with a large number of templates
- Column lock audits now appear consistently
- Fixed issue in config migration where schema field updates were throwing access denied even though user had permission
Compliance Controls
- Entry export with "group by day" fixed
- Entry export with multiple plate designs fixed
- Notebook Entry export and Review Snapshot metadata restored
- Long legal footers display correctly in exports
BioProcess
- Studies field added to workflow task execution
- Closed steps shouldn't reopen improperly
PipeBio
- Fixed issue with importing tsv/csv files
- Fixed "aws import" configuration dialog
- Fixed broken FLASH job
- Fixed PlateLayout chart
- Fix broken AWS Import
- Fixed an issue where a broken Benchling import dialog occurred
- Improve usability of filtering in PipeBio documents