Unverified50% confidenceFactExact time
Speech recognition accuracy has improved from approximately 80% a decade ago to over 95% in standard scenarios today.
1
Sources
50%
Confidence
Long-term
Relevance
9/1/2026
First Seen
Sources
Related Entities
Related Claims
Unverified标准英语场景下，语音转文字（ASR）的转录准确率已超过95%71% similarVerifiedOn the GPQA test, the older advanced voice mode achieved approximately 45% accuracy, while GPT Live mini and the mid-tier version reached 70% to 80%66% similarUnverified说话人分离的主流方案包括基于x-vector嵌入的聚类方法和端到端神经网络方法（如pyannote.audio），标准条件下准确率可达90%以上66% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/833674API
curl https://kongchang.com/api/v1/knowledge/claims/833674MCP
get_claim(id=833674)