Skip to main content

Agora Conversational AI Engine v2.4 Release

×

March 2026 Agora Monthly Digest Meet our new AI agents

×

STT Usage Mismatch Between Reseller RESTful API and Console Due to Outdated SKU Mapping

Updated

Issue Description

When querying Speech-to-Text (STT) usage metrics using the Reseller RESTful API (v2), reported consumption values do not match the aggregate figures displayed in the Reseller Console. The aggregated billing duration or stream count returned by the API appears lower than actual usage.

 

Platform and SDK Context

  • Service / API: Reseller RESTful API (v2) / Agora Speech-to-Text (STT) Service

  • Target Issue: Billing usage data inconsistency between RESTful API response and Reseller Console

     

Error Message

No explicit error message returned; API response shows "retCode": 0 and "msg": "success".

 

Root Cause Analysis

This data discrepancy is caused by outdated or incomplete SKU (Stock Keeping Unit) parameter lists passed in client-side RESTful API requests following backend STT product updates.

  • SKU Model Transition: As the STT architecture evolves (e.g., introduction of higher-concurrency transcription pipelines or new language model tiers), new product SKUs are assigned to track backend usage.

  • Console vs. API Mapping Difference: The Reseller Console and internal billing services automatically incorporate these new SKUs into their baseline query scope. However, RESTful API queries rely on explicit parameters (sku_list or internal product ID filters) provided in the request payload.

  • Silent Telemetry Omission: When an API request specifies a hardcoded or cached SKU list containing only legacy product IDs, the billing server executes the query successfully without throwing an error. However, any STT usage generated under the newly introduced SKUs is omitted from the API response payload, causing a telemetry mismatch.

 

Solution and Resolution Steps

  1. Identify Active and Legacy SKU Mappings

    Audit your API client implementation or billing orchestration service to check if SKU parameters are hardcoded or cached locally.

     

  2. Update Request SKU Parameters

    If you are using cached or hardcoded SKU IDs in your API requests, replace them with the new SKU values corresponding to STT usage.

     

  3. Re-run Query and Validate Data Synchronization

    Execute the updated API request and compare the returned usage_seconds or channel_count metrics against the aggregate reports in the Reseller Console to confirm full data alignment.