← Back to search

allmodels-io/hermes-speech | Plugin to add speech recognition and speech synthesis | GitHub

Kana & Mari's SoundRepos (Japanese) · 2026-09-19 · 2 min
relevance 66 45 words Spotify ↗ Episode page ↗ Audio ↗
Show full episode description
Hermes AgentにAllModels経由の音声合成TTSと音声認識STTを追加するプラグインです。音声ストリーミング、音声検索とプレビュー、モデルや音声の設定、アカウント登録、残高確認、チャージなどをHermesのCLIやTUI、デスクトップアプリから利用できます。 関連: allmodels-io/hermes-speech/音声認識と音声合成を追加するプラグイン/GitHub
✨ Episode Outline — click any point to jump to it in the episode
Problem solved
Hermes AgentにAll Models経由のTTSとSTTを追加依存なしで組み込み、音声選択から利用管理まで一箇所で扱えるようにする。
Benefits
  • TTSとSTTを一つのプラグインでHermes Agentに追加
  • CLI・TUI・デスクトップアプリから音声機能を利用可能
  • 音声ストリーミングと声の検索・プレビューに対応
  • アカウント登録・残高確認・チャージまで同じ場所で管理
  • 追加のPythonパッケージ不要(同梱OpenAI+HTTPXで連携)
Use cases
  • 声を検索してプレビュー試聴してから選ぶワークフロー
  • Hermes同梱のOpenAIクライアントとHTTPXでAll Models APIと連携
  • 音声利用状況の確認とチャージをHermes内で完結
  • PyTestとRuffで開発時の検証を実施
KPIs / results
  • GitHubスター約190
  • Hermes Agent 0.2.0.0以降で動作
  • 追加Pythonパッケージ0個
Tools / build
0:00 / 0:00
🌐 This transcript was automatically translated to English from the original.
Everyone, this is the Kana & Maris SoundRepos plugin that adds speech synthesis and speech recognition to the Hermes Agent. Let's explore how to use voices via All Models. We'll focus on Hermes Speech from All Models-io. We'll use All Models' voice models and voices to add TTS and STT to the Hermes Agent. Not only do we convert text to speech, but we also have the same extension to recognize spoken content. Where can it be used? It can be utilized from Hermes' CLI, TUI, and desktop applications. It also supports audio streaming, audio search, and previews. It's a workflow where you can search, listen to a trial, and then select. Can users configure models and voices? Yes, you can set up models and voices. Account registration, balance checks, charging, and audio usage management are all integrated. You can proceed from searching for voices to checking usage status all in one place. Is an additional package required during installation? No additional Python packages are needed. It connects to the All Models API using OpenAI and HTTPX included with Hermes and operates from Hermes Agent 0.2.0.0 onwards. This setup allows you to try out voice features without increasing dependencies. What do we use for verification during development? We use PyTest and Raf during development. The implementation language is Python, and the license is MIT. It reaches users beyond CLI-only interfaces with different operation screens. It's also easy to understand because you can verify voices in advance with previews. GitHub stars are approximately 190. It's clear that it handles not just audio input/output but also settings and account management. It's clear that recognition and synthesis are not separate tools but can be handled from the entrance of Hermes. It seems we can check the compatibility conditions and try it out. In one sentence, today's certainty is that Hermes Speech is a Python plugin that adds TTS and STT via All Models, voice selection, and usage management to Hermes Agent 0.2.0.0 and later.