NewsEditorialSource: Abestar
Abestar says it removed two translation and speech-recognition models in v1.4.0
Abestar said it removed one text-translation model and one speech-recognition model in its v1.4.0 changelog dated October 13, 2026. The notice describes removal of model choices; the API reference still documents translation and transcription request routes. Reference examples list alternative models for both tasks, but those alternatives were not independently tested. No relevant independent report was found, and live API behavior remains unverified.
Why it matters
Developers using PSSC-V1-251215 or tranc-std-v1 should check code and configuration, query the documented /translations/models and /audio/models discovery routes, and test an explicit supported model. Documentation counts conflict with the examples, so developers should verify available choices before selecting replacements. Abestar’s notice gives no grace period, automatic migration policy or error response for retired IDs, and compatibility for requests that omit a model remains unclear.
Attribution
Sources
- 1AbestarChangelog - Developer Documentation - Abestar APISource published (UTC)
- 2AbestarAPI Reference - Abestar API
Read the original source for full details and context.