Orchestrate your localization workflow with On-demand Fill-ups via API
Tired of waiting for “automatic” triggers to catch up with your specific deployment steps?
While Transifex automation is powerful, complex enterprise workflows often require a specific sequence of events before you want fill-ups to run. For example, you may need to upload existing translations, clean up legacy strings, or complete a custom sync step before triggering Translation Memory, Machine Translation, or AI Fill-ups.
That’s why we’ve introduced On-demand TM, MT, and AI Fill-ups via API. This update gives you more granular, programmatic control over when automation runs for specific resources and languages.
What’s new?
Trigger automation on your own terms
Previously, fill-ups were tied primarily to project-level “on-upload” workflow settings. Now, you can keep those global fill-up settings disabled and use the API to trigger fill-ups when your pipeline is ready.
Instead of relying only on a project-wide approach, you can now:
Target specific resources: Run a fill-up on one specific file without affecting the rest of the project.
Target specific languages: Need to push French live immediately but wait on Japanese? You can trigger fill-ups independently per language.
Supported fill-up combinations
When triggering fill-ups via API, the supported combinations per resource and language are:
- Translation Memory Fill-up
- Translation Memory + AI Fill-up
- Translation Memory + Machine Translation Fill-up
- AI Fill-up
- Machine Translation Fill-up
This means that Translation Memory, AI, and Machine Translation cannot all be enabled together in a single API request.
For workflows that require multiple stages, you can trigger one fill-up first and then trigger another one later for the same resource and language, once the previous fill-up has completed.
For example, a team could first trigger an AI Fill-up and then, after that process is completed, trigger a Machine Translation Fill-up for any remaining untranslated content.
Why this matters
This feature is useful for teams using the Transifex API to manage continuous localization.
More control over timing
No more worrying about a fill-up running before your existing translations have been uploaded or before cleanup steps are finished. You decide when the process starts.
Granular targeting
Trigger fill-ups for specific resources and languages instead of relying only on project-wide automation.
Flexible automation paths
Use supported combinations such as TM, TM + AI, TM + MT, AI, or MT depending on your workflow and quality requirements.
Reduced risk of overlapping automation
API-triggered fill-ups should be coordinated carefully. If you need to run more than one fill-up for the same resource and language, make sure the first one has completed before triggering the next one.
How it works
To use this functionality, automated fill-ups must be disabled in the Project Workflow settings. API requests are blocked if any fill-up option is enabled in the project workflow settings.
Once project-level fill-ups are disabled, you can send a POST request to the resource_translations_async_fillups endpoint with the desired supported fill-up configuration for a specific resource and language.
Possible parameters may include:
translation_memory_fillup: true/false
machine_translation_fillup: true/false
ai_fillup: true/false
However, not every combination is supported in a single request. Supported combinations are:
TM
TM + AI
TM + MT
AI
MT
Pro Tip
Use this to create a “Translation Warm-up” script. Every time a developer pushes a new feature branch, your script can trigger a supported fill-up combination for that branch’s resource and target language, such as TM + AI or TM + MT, depending on your workflow.
This helps ensure that key strings are populated early while still giving your team control over when and how automation runs.
Conclusion
By moving from a project-wide trigger to targeted API execution, Transifex gives developers and localization managers more precision for complex localization environments.
You stay in control of the workflow, the quality, and the timing — deciding exactly when fill-ups should run, for which resources, and for which languages.
Check out the full technical details in our Developer Hub and let us know how you’re planning to orchestrate your new automated pipelines!