[ 25.08.31 ] August API update monthly summary

Summary

Several customer-requested features were added, including expanded Omni sending, automatic number normalization, and long-term MMS storage.

At a glance

• A per-recipient ref field for Omni sending
• Automatic normalization of '-' inside numbers
• MMS file retention extended from 7 days to 1 year
• responseMethod / timeout options for Alimtalk

What changed?

A per-number ref field for sending

On a broadcast send you can set a separate reference value (ref) for each recipient number, which makes downstream processing and status tracking more precise.

2. Automatic normalization of '-' inside numbers

Even if you pass numbers with hyphens, they are converted to digits-only automatically, with no normalization step on your side.

3. Long-term MMS file storage (7 days → 1 year)

MMS images and attachments are kept for up to a year, which brings the policy in line with the existing RCS and Kakao image retention rules.

4. Choose how Alimtalk results are received

Set responseMethod(polling/push) and timeout to pick the result delivery method that fits your service.

Why it matters

• A per-number ref field lets enterprise customers sending large volumes map and analyze each message precisely, which raises operational efficiency.
• Automatic number normalization reduces input errors, which lowers the failure rate and makes development and operations easier.
• **Long-term MMS storage (1 year)** meets the long-term log and history analysis needs of enterprise customers, improving settlement, security, and audit response.
• Choosing the Alimtalk result method lets you pick polling or push based on your architecture, which adds a lot of development flexibility.

How to use it

• Set the ref field per number when you send, and matching individual messages during status lookup and report analysis becomes straightforward.
• For Alimtalk, choose polling or push with responseMethod and tune the timeout value to suit your service.

Things to note

• Setting timeout too low can cause a timeout before the channel responds, so pick a value that fits your architecture.