Coming soon

About This Software

Press the hotkey, speak, and text appears at your cursor

ORI Voice Input is a real-time voice dictation tool for Windows. Press the global hotkey and start speaking — recognized text appears character by character right at your cursor, in any input field, editor, or chat window. It even works in games.

Key Features

• Live typing: Text appears as you speak; when the engine revises earlier words, it backspaces and rewrites automatically. You can also switch to committing a full sentence at once.

• No swallowed first words: An always-warm microphone plus a ~400 ms pre-roll buffer means the beginning of every take is never lost.

• Noise-robust segmentation: Sentence breaks are detected from voice energy, with automatic ambient-noise calibration every time recording starts.

• Instant take-back: Press Delete or left-click to interrupt; everything not yet committed in the current take is discarded.

• Push-to-talk: Hold the hotkey to speak; release to stop and commit immediately.

• Automatic wrap-up: Recording stops automatically after 10 seconds of silence, with a maximum of 60 seconds per session (both adjustable).

• Game mode: Works with games' self-drawn input boxes via system-level clipboard pasting.

What's Included in the Steam Edition

• Built-in offline recognition engine and dependencies — ready out of the box, zero configuration

• Steam Workshop: community content such as feature packs, themes, fonts, phrases, and prompts

• Steam Cloud: config files and history sync to the cloud by default

• Steam Stats: only cumulative recording time, recognized character count, and session count — never any text or audio

• Exclusive pixel-art theme

Built-in Offline Engine

The Steam Edition ships with a local recognition engine and all dependencies, ready right after installation: no account registration, no API keys, fully offline, and no usage fees.

Please note: the built-in offline engine currently supports Chinese speech only. For high-accuracy English dictation, a cloud engine is required (you provide your own API keys).

Interface & System Integration

• Floating status bar: four built-in themes (Light / Dark / Minimal / Pixel), with adjustable font, position, size, opacity, and auto-hide.

• History: recognized content can be recorded, with a standalone window to search the full history.

• Automatically pauses background media or mutes system audio while recording, and restores it when you stop.

Text Injection & Compatibility

Two delivery methods are supported: clipboard paste and per-character simulated keystrokes, with automatic protection and restoration of your clipboard contents.

Game mode always uses clipboard injection to work with self-drawn input boxes. When the target app runs as administrator, you can elevate with one click or always launch as administrator.

Voice Commands

Once enabled, local commands such as "listen: select all" or "listen: undo" consume no LLM calls; custom phrases insert preset text with a single command.

Feature Packs Add High-Accuracy Cloud Engines

• Alibaba Cloud Bailian (Qwen speech recognition)

• Tencent Cloud real-time speech recognition

• iFLYTEK real-time transcription (LLM edition / standard edition)

• Volcano Engine Doubao streaming speech recognition

Switch engines anytime from the right-click menu — no restart needed.

Cloud services are activated and billed directly between you and each provider. All providers offer free trial quotas; this software does not collect any fees on their behalf.

Feature Packs Add AI Features

• AI correction: After recognition, automatically fixes homophone typos, completes punctuation, and removes filler words; the prompt is customizable and compatible with any OpenAI-protocol endpoint.

• AI voice commands: Say "help me polish / translate / expand" to run AI processing on the selected text.

• Live voice conversation: Supports the Doubao end-to-end speech LLM, Alibaba Cloud Qwen-Audio, and self-hosted Hermes Agent; barge in anytime while the AI is speaking to take over, with bubble captions shown on the floating bar.

Privacy

With the built-in offline engine, audio and recognition results are processed entirely on your machine. When using cloud engines or AI correction, the audio or text is sent directly to the provider you configured.

AI Generated Content Disclosure

The developers describe how their game uses AI Generated Content like this:

During use, the speech recognition and AI features connect to cloud services configured by you (Alibaba Cloud / Tencent Cloud / iFlytek / Volcano Engine speech recognition, or OpenAI-compatible LLM endpoints), and the resulting text is generated in real time by the selected service.