Unverified50% confidenceBenchmarkExact time
MedAstra's pure speech retriever achieves 87.6 Recall@10, the hybrid configuration achieves 90.4, and transcription-first dense retrieval achieves 81.0.
1
Sources
50%
Confidence
Long-term
Relevance
11/3/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedWhen streaming audio back in Flask, chunk_size=8192 is a compromise value: too small increases request overhead, while too large raises the time to first byte; the returned mimetype must be audio/mpeg, otherwise the <audio> tag may not parse correctly.62% similarUnverifiedElevenLabs launched the real-time speech-to-speech models Max and Mimi; Max scored 98.8% on Big Bench Audio and begins speaking in about 0.57 seconds.60% similarUnverifiedWhisper audio is first converted into an 80-dimensional mel spectrogram, downsampled by convolutional layers, fed into the encoder to generate a feature sequence, and then decoded by the decoder to output text token by token in an autoregressive manner60% similarUnverifiedThis version supports transcription model IDs and adds vadThreshold (voice activity detection threshold) and opus audio format support.60% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1129049API
curl https://kongchang.com/api/v1/knowledge/claims/1129049MCP
get_claim(id=1129049)