Smoke Test

Run

Default target: faster_whisper + local. Use TARGET_LANGUAGES=kk for first smoke; cloud providers are fallback/legacy.

Lesson Audio source STT provider Translator Target languages

Advanced / Legacy fallback providers

Legacy/cloud fallback, not current MVP: STT mock, ElevenLabs, Azure, Cartesia; translation mock and Azure. Use API overrides for fallback-only tests.

Glossary Enable glossary

Current MVP: Russian → Kazakh only

STT target: faster-whisper / ru / auto-detect off
Translation kk: Tilmash ru→kk
Translation uz: disabled for MVP
Translation zh-Hans: disabled for MVP
M2M100: disabled for MVP

Legacy/fallback cloud providers

mock STT: ready
ElevenLabs: missing ELEVENLABS_API_KEY
Azure STT: missing AZURE_SPEECH_KEY, AZURE_SPEECH_REGION
Cartesia STT: missing CARTESIA_API_KEY
mock translator: ready
Azure: missing AZURE_TRANSLATOR_KEY

Live Results

status: idle
smoke_test_id: none
lesson_id: none
raw original
normalized
kk
uz
zh-Hans

Latency

first partial: 0 ms
STT final: 0 ms
translation: 0 ms
server total: 0 ms
client receive: 0 ms
estimated usage: not calculated

Audio Streaming

mode: none
audio duration: 0 ms
chunks: 0
chunk size: 0 ms
send elapsed: 0 ms
realtime factor: 0

Events

Latest Smoke Tests

No smoke tests yet.