Skip to main content

Smartling MCP Server Tools

The Smartling MCP server lets AI tools such as ChatGPT and Claude work with your Smartling account through chat. Your AI tool selects the tools it needs based on your prompt. To connect your AI tool, see Smartling MCP Server: Connect Your AI Chat Tools to Smartling.


MCP tools are grouped into toolsets, such as jobs, strings, and glossary, so you can enable just the tools relevant to your work and conserve context-window capacity. By default, your AI tool sees only a curated set of about 78 tools called core. To see more tools, select additional toolsets in your server URL or request headers. For example, this URL shows only job and string tools:

https://mcp.smartling.com/mcp?toolsets=jobs,strings

Use toolsets=all to show every tool. See Configure MCP toolsets for your AI tool for all options.


Tools are listed by toolset below. Each section shows its toolset name, and tools marked (core) are available by default. Tools without the tag need their toolset selected, unless the toolset is marked "always available."

Info: Toolset settings only apply if you configure the Smartling MCP server yourself, such as in Claude Code, Cursor, or Visual Studio Code, or through a custom connector. If you add Smartling MCP from the ChatGPT or Claude directory, only the core tools are available. To access additional tools, ask your organization admin to configure a custom connector with the tools you need.


Configure MCP toolsets for your AI tool

AI tools limit how many tools they can load at once (commonly about 128). To stay under that limit as the Smartling MCP server grows, the server shows only the core set unless you ask for more. Toolsets also help you keep your AI tool focused. For example, if you only work with jobs and strings, selecting those two toolsets means your AI tool loads fewer tools and saves space in your context window.

Note: Toolsets are not a security control, and they don't limit what you are authorized to do. Your ability to use certain MCP tools is determined by how you authenticate and your Smartling user role.


Settings

Each setting can be sent as an HTTP header or a URL query parameter:

Purpose

Header

Query param

Notes

Select toolsets

X-MCP-Toolsets

toolsets

Comma-separated toolset names (also accepts the core and all values below)

Force-include tools

X-MCP-Tools-Add

add_tools

Comma-separated tool names, added regardless of toolset selection

Force-exclude tools

X-MCP-Tools-Omit

omit_tools

Comma-separated tool names, excluded regardless of toolset selection or the add list

You can also use two special values anywhere a toolset name is accepted, in addition to real toolset names:

Value

Effect

core

The curated default set (about 78 tools). Used automatically when no toolset is selected. Combine it with real toolset names to extend it, for example toolsets=core,estimates.

all

Every tool across every toolset, bypassing the core filter entirely. Wins outright over any other entries present alongside it.


How the settings combine

  • Header vs. query parameter: If you set both, the header wins, as long as it contains at least one valid value. If the header is present but blank (for example, a proxy adds an empty header), the server uses the query parameter instead. This applies separately to toolsets, added tools, and excluded tools.

  • Add vs. omit: Omit always wins. A tool in both add_tools and omit_tools is removed.

  • Empty or unrecognized selection: If you don't select any toolsets, or none of the names are recognized, the server uses core, not all tools.

  • Spelling: Names are not case-sensitive. The server ignores names it doesn't recognize without an error, so if a tool is missing, check the spelling.

  • Named toolsets: When you name a toolset, you get every tool in it, whether or not the tool is tagged (core). Only the core value filters by tool.

  • Always available: The skills and file-upload toolsets are always available, whatever you select.


Examples

Default (core toolset, no selector sent):

https://mcp.smartling.com/mcp

Every tool:

https://mcp.smartling.com/mcp?toolsets=all

The default tools plus the estimates toolset:

https://mcp.smartling.com/mcp?toolsets=core,estimates

Select specific toolsets, using a query parameter:

https://mcp.smartling.com/mcp?toolsets=jobs,strings

Select specific toolsets, using a header:

X-MCP-Toolsets: jobs,strings

Add one extra tool on top of a toolset:

https://mcp.smartling.com/mcp?toolsets=jobs&add_tools=smartling_list_glossaries

Omit one specific tool from a toolset:

https://mcp.smartling.com/mcp?toolsets=jobs&omit_tools=smartling_manage_job_status

Blank header falls back to the query parameter:

Header: X-MCP-Toolsets:  (empty) URL:    ?toolsets=strings Result: "strings" toolset is used (query param applies since the header resolved to nothing)

Toolset names

jobs, strings, instant-translation, glossary, issues, lqa, styleguide, people, files, file-upload, job-batches, integrations, linguistic-package, project, reports, estimates, skills, tags, translation-memory, vendors, workflow


Tools by toolset

Job management tools

Toolset: jobs

Create, update, authorize, and track translation jobs and the files and strings in them.

  • smartling_list_jobs (core): Lists translation jobs from a specific project, with optional filtering by name and status.

  • smartling_create_job (core): Creates a new Smartling job with full control over custom fields, reference numbers, and due dates. See smartling_get_or_create_translation_job for a simpler alternative.

  • smartling_get_job (core): Retrieves detailed information about a specific job by UID, including priority, source files, and metadata.

  • smartling_update_job (core): Updates a job's metadata (name, description, due date, reference number, callback configuration, custom fields). Any optional field you leave out is cleared.

  • smartling_add_strings_to_job (core): Adds strings to an existing job by their hashcodes, with optional target locale and move settings.

  • smartling_remove_strings_from_job (core): Removes strings from an existing job by their hashcodes, for specific or all target locales.

  • smartling_authorize_job (core): Authorizes a job to begin translation work, moving it from preparation to active status.

  • smartling_get_job_history (core): Retrieves the audit history of a job (creation, status changes, content, due date, and priority changes), with pagination through offset and limit.

  • smartling_get_job_progress (core): Gets per-locale, per-workflow-step progress for a translation job, optionally scoped to a single locale or file.

  • smartling_list_job_files (core): Lists the source files attached to a Smartling job, with pagination.

  • smartling_list_file_jobs (core): Lists the translation jobs a specific file is involved in, flagging any unassigned content.

  • smartling_manage_job_file (core): Adds or removes a file from a Smartling job.

  • smartling_manage_job_status (core): Changes a job's lifecycle status (cancel, close, or delete).

Job batch tools

Toolset: job-batches

Upload multiple files and attach them to a translation job in one flow.

  • smartling_create_job_batch (core): Creates a batch to upload multiple files and attach them to a translation job in one coordinated flow.

  • smartling_get_or_create_translation_job (core): Gets an existing translation job by name template, or creates one. A simpler alternative to smartling_create_job for batch and file-upload workflows.

  • smartling_process_job_batch_action (core): Registers or cancels a single file within a job batch.

  • smartling_upload_file_to_job_batch (core): Uploads a file's content into a job batch. Supply the file with an upload_id from smartling_request_upload_url, or send the file content directly if your client can't run commands.

  • smartling_get_job_batch_status (core): Retrieves detailed batch status, including per-file, per-locale progress.

  • smartling_list_job_batches (core): Lists job batches for a project, with optional filtering by translation job or status.

Machine translation tools

Toolset: instant-translation

Translate text and files with machine translation, without creating a job.

  • smartling_machine_translate_text (core): Translates text items between languages, with automatic language detection.

  • smartling_machine_translate_file (core): Translates entire files in various formats. Supply the file with an upload_id from smartling_request_upload_url, or send the file content directly if your client can't run commands. Returns a download link (valid for 15 minutes) instead of the file.

File upload tools

Toolset: file-upload (always available)

Supply files to other tools from your computer.

  • smartling_request_upload_url: Returns a short-lived upload link for a file. This is the default way to supply a file for clients that can run commands on your computer, whatever the file size.

String tools

Toolset: strings

Search for and create strings in a project.

  • smartling_search_strings (core): Searches for strings in a Smartling project using various filters (status, locales, source keywords).

  • smartling_create_strings (core): Creates strings in a Smartling project and waits until creation finishes.

String tagging tools

Toolset: tags

Add and remove tags on strings.

  • smartling_add_tags_to_strings (core): Adds tags to strings for categorization and filtering (up to 100 tags and 1,000 strings per request).

  • smartling_remove_tags_from_strings: Removes specific tags from strings (up to 100 tags and 1,000 strings per request).

  • smartling_remove_all_tags_from_strings: Removes all tags from strings for complete cleanup (up to 1,000 strings per request).

File management tools

Toolset: files

Check file translation status, find recent files, and download translated content.

  • smartling_get_file_status (core): Retrieves a file's translation status for every locale, and its last-modified date per locale.

  • smartling_search_recent_files (core): Searches recently uploaded source files or recently published translated files.

  • smartling_download_translated_file (core): Downloads translated content for a file: a single file and locale, a zip of all locales, or a zip of multiple files across locales. Returns a download link (valid for 15 minutes) instead of the file.

  • smartling_manage_file: Manages a file's lifecycle in a project (supports deleting or renaming a file).

Glossary management tools

Toolset: glossary

Manage glossaries, glossary entries, and labels, and match terms against text.

  • smartling_list_glossaries (core): Lists glossaries, with advanced filtering (query, state, locale, UIDs).

  • smartling_get_glossary (core): Retrieves complete glossary configuration by UID.

  • smartling_match_glossary_terms (core): Matches source and translation text against glossary terms.

  • smartling_export_glossary (core): Exports a glossary's terms as CSV, XLSX, or TBX. Returns a download link (valid for 15 minutes) instead of the file.

  • smartling_manage_glossary: Creates a new glossary or updates an existing one.

  • smartling_manage_glossaries_archive_state: Archives or restores one or more glossaries in bulk.

  • smartling_get_glossary_entry (core): Retrieves a single glossary entry by its UID.

  • smartling_search_glossary_entries (core): Searches entries within a glossary by a set of filters.

  • smartling_manage_glossary_entry (core): Creates a new glossary entry or updates an existing one.

  • smartling_manage_glossary_entries_bulk: Archives, restores, removes, or adds or removes labels on glossary entries matching a filter, in bulk.

  • smartling_list_glossary_labels: Lists all glossary labels in an account.

  • smartling_manage_label: Creates, updates, or deletes a glossary label.

  • smartling_manage_glossary_import (core): Imports glossary entries from a file (CSV, XML, or XLSX) through a start, status, and confirm workflow.

  • smartling_authorize_entries_for_translation: Authorizes glossary entries matching a filter for translation into one or more locales through existing project workflows.

Style guide tools

Toolset: styleguide

Create and manage style guides, their locales, and their attachments.

  • smartling_list_styleguides (core): Lists style guides, with optional name filtering.

  • smartling_get_styleguide (core): Retrieves detailed style guide configuration by UID.

  • smartling_create_styleguide: Creates a new style guide, either PLAINTEXT or SMARTLING-type (recommended).

  • smartling_update_styleguide: Updates an existing style guide's content and settings.

  • smartling_delete_styleguide: Permanently deletes a style guide.

  • smartling_set_styleguide_locales: Replaces the full set of target locales bound to a style guide.

  • smartling_list_styleguide_locales: Lists the target locales currently bound to a style guide.

  • smartling_add_styleguide_attachment: Adds a link or file attachment to a style guide.

  • smartling_list_styleguide_attachments: Lists the attachments on a style guide.

  • smartling_remove_styleguide_attachment: Removes an attachment from a style guide.

  • smartling_download_styleguide_attachment: Downloads a file attachment from a style guide. Returns a download link (valid for 15 minutes) instead of the file.

  • smartling_get_styleguide_permissions: Gets your style guide permissions.

Linguistic package tools

Toolset: linguistic-package

View and manage linguistic packages, which bundle glossaries and style guides.

  • smartling_list_linguistic_packages (core): Lists all linguistic packages from the current account.

  • smartling_get_linguistic_package (core): Retrieves linguistic package configuration by UID.

  • smartling_create_linguistic_package: Creates a new linguistic package.

  • smartling_update_linguistic_package: Updates an existing linguistic package. Replaces the full configuration.

  • smartling_delete_linguistic_package: Deletes a linguistic package.

  • smartling_get_linguistic_packages_by_glossaries: Finds the linguistic packages that reference each of the given glossaries.

Translation memory tools

Toolset: translation-memory

Search, manage, import, and export translation memories.

  • smartling_search_translation_memory (core): Searches translation memory for similar translations in a specific project and locale.

  • smartling_search_translation_memories (core): Lists the translation memories in an account, optionally filtered by name, source locale, or translation quality.

  • smartling_get_translation_memory (core): Retrieves a single translation memory by UID.

  • smartling_create_translation_memory: Creates a new, empty translation memory.

  • smartling_update_translation_memory: Renames or re-describes an existing translation memory.

  • smartling_delete_translation_memory: Deletes a translation memory and all of its strings.

  • smartling_import_tmx_file: Imports a TMX file's translation units into an existing translation memory. Supply the file with an upload_id from smartling_request_upload_url, or send the file content directly if your client can't run commands.

  • smartling_export_tmx_file: Searches a translation memory's strings by filter criteria and exports the matches as a TMX file. Returns a download link (valid for 15 minutes) instead of the file.

Project management tools

Toolset: project

List projects and get project details.

  • smartling_list_projects (core): Lists Smartling projects, with filtering options (name, type codes, archived status).

  • smartling_get_project (core): Retrieves detailed project information by UID.

Workflow tools

Toolset: workflow

List the workflows configured in a project.

  • smartling_list_workflows (core): Lists workflows from a specific Smartling project, with optional filtering.

Report tools

Toolset: reports

Get word count and string change reports.

  • smartling_get_word_count_report (core): Gets a word count report (projects, jobs, agencies, locales, workflow steps, fuzzy matching) from Smartling in CSV format. Returns a download link (valid for 15 minutes) instead of the file.

  • smartling_get_string_changes_report: Gets a report tracking how translations were modified over time, including incoming and outgoing translation text, edit distances, translator information, workflow details, and timestamps.

  • smartling_get_workflow_step_types (core): Gets the available workflow step types, for use in filtering other report tools.

Estimate tools

Toolset: estimates (no tools in core)

Generate, view, and download fuzzy match and cost estimate reports for jobs.

  • smartling_list_job_fuzzy_estimate_reports: Lists fuzzy match estimate reports previously generated for a job, with filtering and pagination.

  • smartling_generate_job_fuzzy_estimate_report: Generates a new fuzzy match estimate report for a job. Report generation is asynchronous, so check progress with smartling_get_estimate_report_status.

  • smartling_list_job_cost_estimate_reports: Lists cost estimate reports previously generated for a job, with filtering and pagination.

  • smartling_generate_job_cost_estimate_report: Generates a new cost estimate report for a job. Report generation is asynchronous, so check progress with smartling_get_estimate_report_status.

  • smartling_get_estimate_report: Gets the full contents of a fuzzy match or cost estimate report.

  • smartling_delete_estimate_report: Deletes a fuzzy match or cost estimate report.

  • smartling_get_estimate_report_status: Gets the generation status and completion percentage of an estimate report.

  • smartling_update_estimate_report_tags: Replaces the tags on an estimate report.

  • smartling_download_estimate_report: Downloads an estimate report as a CSV file. Returns a download link (valid for 15 minutes) instead of the file.

People management tools

Toolset: people

List accounts, users, and roles, and invite or remove users.

  • smartling_get_available_accounts_for_current_user (core): Gets available accounts for the current user. Use this before other Smartling tools that need an accountUid parameter.

  • smartling_list_users (core): Lists users from the current Smartling account, with advanced filtering options (keyword, roles, projects, status, pagination).

  • smartling_get_current_user_roles_info (core): Gets your top native role, your currently active role, and the roles you can switch to.

  • smartling_list_current_user_projects (core): Lists the projects you can access within an account.

  • smartling_get_user_allowed_actions: Gets the actions you are allowed to perform on specific users in an account.

  • smartling_remove_users: Removes users from a Smartling account, revoking their access entirely.

  • smartling_list_invites: Lists pending invites for a Smartling account, with filtering, sorting, and pagination.

  • smartling_get_invite_link: Gets the shareable invitation link URL for a pending invite.

  • smartling_manage_invites: Resends or cancels pending invites, identified by invite UIDs or user UIDs.

  • smartling_get_account_user_status: Checks whether email addresses already belong to a user, account user, or pending invite.

  • smartling_invite_users: Invites one or more new users to a Smartling account with account-level, project-level, and workflow-level roles.

Vendor tools

Toolset: vendors

View content assignments for translators and vendors.

  • smartling_list_content_assignments (core): Lists content assignments for the current account (project, locale, workflow step, translation job, assignee, due dates, word count).

Issue tools

Toolset: issues

Create, update, comment on, and search string issues.

String issues capture questions, clarification requests, and quality flags raised against source or translated strings.

Issue lifecycle

  • smartling_create_issue (core): Creates a new issue on a string with type, sub-type, severity, and assignee.

  • smartling_get_issue (core): Retrieves full details of one issue by UID.

  • smartling_edit_issue_text: Edits the body text of an existing issue (maximum 4,000 characters).

  • smartling_open_or_close_issue (core): Changes an issue's state to OPENED or RESOLVED.

  • smartling_update_issue_answered (core): Toggles the answered flag.

  • smartling_update_issue_assignee: Assigns an issue to a specific user.

  • smartling_delete_issue_assignee: Removes the current assignee.

  • smartling_update_issue_severity_level: Changes severity (LOW, MEDIUM, HIGH).

  • smartling_update_issue_type: Changes type and sub-type, and optionally narrows to a locale.

Comments

  • smartling_list_issue_comments (core): Lists all comments on an issue.

  • smartling_get_issue_comment: Retrieves one comment by UID.

  • smartling_add_issue_comment (core): Posts a new comment (maximum 4,000 characters).

  • smartling_update_issue_comment: Edits an existing comment.

  • smartling_delete_issue_comment: Deletes a comment.

Search and count

  • smartling_find_project_issues (core): Searches issues in one project, with pagination and sorting.

  • smartling_count_project_issues: Counts issues in one project matching the filter.

  • smartling_find_account_issues (core): Searches issues across all projects in your account.

  • smartling_count_account_issues: Counts issues across all projects in your account.

LQA tools

Toolset: lqa

Run automated quality evaluations, record and search quality findings, and get MQM scores.

  • smartling_execute_lqa_evaluation (core): Runs an automated Linguistic Quality Assurance (LQA) evaluation on a single translated string and returns the detected quality errors (category, severity).

  • smartling_list_lqa_schemas (core): Lists LQA schemas configured for an account.

  • smartling_get_lqa_schema_topology (core): Gets the category, severity, and error-type breakdown for one LQA schema (resolves human-readable names to UIDs).

  • smartling_record_lqa_error (core): Records a manual LQA finding on a translated string.

  • smartling_update_lqa_error (core): Updates an existing recorded finding's error type, severity, or comment.

  • smartling_delete_lqa_error: Deletes a recorded finding.

  • smartling_mark_lqa_no_errors: Marks a translated string as reviewed with no errors found.

  • smartling_list_lqa_errors_per_string (core): Reads back findings recorded on one specific string.

  • smartling_find_account_lqa_errors (core): Account-wide filtered search across recorded findings.

  • smartling_export_account_lqa_errors: Exports account-wide findings as a CSV, using the same filters as smartling_find_account_lqa_errors. Returns a download link (valid for 15 minutes) instead of the file.

  • smartling_get_mqm_statistics (core): Gets MQM (Multidimensional Quality Metrics) quality scores for one or more jobs, where available.

Skill tools

Toolset: skills (always available)

Skills are bundles of instructional text the server can hand an AI model on demand, for capabilities that don't need a dedicated tool.

  • smartling_list_skills: Lists available skills, grouped by category, with an optional category filter.

  • smartling_load_skill: Loads the full instructions for one skill by id, for the AI model to follow directly.

Integration and connector tools

Toolset: integrations

Search connector content and request human or machine translation for it.

Connectors pull content from a third-party system into Smartling for search and translation. Call smartling_integration_list_available first to see which connectors exist on an account, then smartling_integration_get_search_context before smartling_integration_search_assets to learn valid filter values.

  • smartling_integration_list_available (core): Lists connectors available on an account and their normalized integration ID.

  • smartling_integration_get_search_context (core): Fetches connection status, valid filter values (asset types, submission statuses, connector-specific filters), and locale mappings for a connector and project.

  • smartling_integration_search_assets (core): Searches a connector's content for assets to translate.

  • smartling_integration_request_tms_translation (core): Submits selected assets to an existing Smartling job for human translation.

  • smartling_integration_request_instant_translation (core): Requests on-the-fly machine translation for selected assets.

  • smartling_integration_search_instant_translation_requests (core): Lists and filters past instant translation requests.

  • smartling_integration_get_instant_translation_request (core): Gets per-asset, per-locale results of one instant translation request.

Downloading generated files

Eight tools return a download link instead of the file itself: smartling_download_translated_file, smartling_export_glossary, smartling_machine_translate_file, smartling_export_account_lqa_errors, smartling_get_word_count_report, smartling_download_styleguide_attachment, smartling_download_estimate_report, and smartling_export_tmx_file. The link is valid for 15 minutes. After that, run the tool again to get a new one. This keeps large files, such as zipped translations and exports, out of the conversation with your AI tool, so open the link directly to get the content.

Did this answer your question?