llama.cpp b12238 新增 OpenAI 兼容语音端点
原标题：b12238
AI 摘要
llama.cpp 发布 b12238 版本，其 server 组件新增了与 OpenAI API 兼容的 /v1/audio/speech 端点，用于文本转语音（TTS）输出。该 PR 同时补充了端到端 pytest 测试，并统一了 WAV 写入与 pcm_f32_to_s16 的舍入行为，使 /tts 与 /v1/audio/speech 共用 parse_tts_body 解析请求体。此次更新让本地推理服务能直接对接 OpenAI 语音接口生态，并随版本提供 macOS、Linux、Windows、Android 等多平台预编译包。
正文节选
<details open> server: add oai-compat /v1/audio/speech endpoint (#31763) * server: add oai-compat /v1/audio/speech endpoint * add e2e pytest * server: share the s16 rounding, type speech form fields The WAV writer of the audio helper truncates while pcm_f32_to_s16 rounds, so wav and pcm differ by one LSB. Both now round. pcm_f32_to_s16 takes a pointer and a count and converts the TTS bytes without an intermediate copy. /tts and /v1/audio/speech parse their body with parse_tts_body, so mul