Speechall command-line tool for fast speech-to-text transcription using multiple providers — Speechall command-line 工具 for fast speech-to-text transcription using multiple 提供者s

speechall-命令行工具

命令行工具 for speech-to-text transcription via the Speechall API. Supports multiple 提供者s (OpenAI, Deepgram, AssemblyAI, Google, Gemini, Groq, ElevenLabs, Cloudflare, and more).

安装ation Homebrew (macOS and Linux) brew 安装 Speechall/tap/speechall

Without Homebrew: 下载 the binary for your 平台 from https://github.com/Speechall/speechall-命令行工具/releases and place it on your PATH.

验证 speechall --version

Authentication

An API key is required. Provide it via 环境 variable (preferred) or flag:

导出 SPEECHALL_API_KEY="your-key-here" # or speechall --API-key "your-key-here" audio.wav

The user can 创建 an API key on https://speechall.com/console/API-keys

Commands transcribe (default)

Transcribe an audio or video file. This is the default subcommand — speechall audio.wav is equivalent to speechall transcribe audio.wav.

speechall [options]

Options:

Flag Description Default --模型 <提供者.模型> STT 模型 identifier openAI.gpt-4o-mini-transcribe --language Language code (e.g. en, tr, de) API default (auto-检测) --输出-格式化 <格式化> 输出 格式化 (text, json, verbose_json, srt, vtt) API default --diarization Enable speaker diarization off --speakers-expected Expected number of speakers (use with --diarization) — --no-punctuation Disable automatic punctuation — --temperature <0.0-1.0> 模型 temperature — --initial-prompt Text prompt to 图形界面de 模型 style — --custom-vocabulary Terms to boost recognition (repeatable) — --rule设置-id Replacement rule设置 UUID — --API-key API key (overrides SPEECHALL_API_KEY env var) —

Examples:

# Basic transcription speechall interview.mp3

# Specific 模型 and language speechall call.wav --模型 deepgram.nova-2 --language en

# Speaker diarization with SRT 输出 speechall meeting.wav --diarization --speakers-expected 3 --输出-格式化 srt

# Custom vocabulary for domAIn-specific terms speechall medical.wav --custom-vocabulary "myocardial" --custom-vocabulary "infarction"

# Transcribe a video file (macOS 提取s audio automatically) speechall presentation.mp4

模型s

列出 avAIlable speech-to-text 模型s. 输出s JSON to stdout. 过滤器s combine with AND 记录ic.

speechall 模型s [options]

过滤器 flags:

Flag Description --提供者 过滤器 by 提供者 (e.g. openAI, deepgram) --language 过滤器 by supported language (tr matches tr, tr-TR, tr-CY) --diarization Only 模型s supporting speaker diarization --srt Only 模型s supporting SRT 输出 --vtt Only 模型s supporting VTT 输出 --punctuation Only 模型s supporting automatic punctuation --流able Only 模型s supporting real-time 流ing --vocabulary Only 模型s supporting custom vocabulary

Examples:

# 列出 all avAIlable 模型s speechall 模型s

# 模型s from a specific 提供者 speechall 模型s --提供者 deepgram

# 模型s that support Turkish and diarization speechall 模型s --language tr --diarization

# Pipe to jq for specific fields speechall 模型s --提供者 openAI | jq '.[].identifier'

Tips On macOS, video files (.mp4, .mov, etc.) are automatically converted to audio before 上传. On Linux, pass audio files directly (.wav, .mp3, .m4a, .flac, etc.). 输出 goes to stdout. Redirect to save: speechall audio.wav > transcript.txt Errors go to stderr, so piping stdout is safe. 运行 speechall --help, speechall transcribe --help, or speechall 模型s --help to see all valid enum values for 模型 identifiers, language codes, and 输出 格式化s.