TL;DR
audio request property type/format has changed for the /accounts/{account_id}/ai/run/@cf/openai/whisper-large-v3-turbo endpoint./accounts/{account_id}/ai/run/@cf/openai/whisper-large-v3-turbo endpoint, including beam_size, compression_ratio_threshold, condition_on_previous_text, hallucination_silence_threshold, log_prob_threshold, and no_speech_threshold.anyOf list for the audio request property has been updated to include 'subschema #1, subschema #2'./accounts/{account_id}/ai/run/@cf/zai-org/glm-4.7-flash./accounts/{account_id}/ai/run/{model_name} endpoint.Editor's Note:
We've observed that this update is part of a series of enhancements to the AI-related endpoints, reflecting a broader trend towards more flexible and customizable AI models. As we review these changes, our tip for developers is to carefully examine the updated audio request property and its new optional properties to ensure compatibility with existing code. This is not a routine maintenance release, but rather a significant shift in the capabilities of the AI endpoints.
What Changed
audio request property type/format has changed from 'string'/' ' to ' '/' ' for the /accounts/{account_id}/ai/run/@cf/openai/whisper-large-v3-turbo endpoint./accounts/{account_id}/ai/run/@cf/openai/whisper-large-v3-turbo endpoint, including beam_size, compression_ratio_threshold, condition_on_previous_text, hallucination_silence_threshold, log_prob_threshold, and no_speech_threshold./accounts/{account_id}/ai/run/@cf/zai-org/glm-4.7-flash./accounts/{account_id}/ai/run/{model_name} endpoint.Who's Impacted
If you are currently using the /accounts/{account_id}/ai/run/@cf/openai/whisper-large-v3-turbo endpoint, you will need to update your code to reflect the changes to the audio request property. If you are using the /accounts/{account_id}/ai/run/{model_name} endpoint, you may need to adjust your code to handle the widened list-of-types for the request property '/oneOf[subschema #12: Image Text To Text]/oneOf[subschema #2]/messages/items/content'. If you are planning to use the new /accounts/{account_id}/ai/run/@cf/zai-org/glm-4.7-flash endpoint, you should review the updated documentation to ensure compatibility with your existing code.
Action Checklist
audio request property for the /accounts/{account_id}/ai/run/@cf/openai/whisper-large-v3-turbo endpoint.audio request property.audio request property./accounts/{account_id}/ai/run/@cf/openai/whisper-large-v3-turbo endpoint./accounts/{account_id}/ai/run/{model_name} endpoint./accounts/{account_id}/ai/run/@cf/zai-org/glm-4.7-flash endpoint documentation.Verification
The information in this update brief is based on the following sources:
Last verified at: 2026-02-20T02:42:26.591811+00:00