OpenAI Web Console

A browser-direct console for OpenAI's API: transcribe audio with Whisper or GPT-4o Transcribe, translate audio into English, generate speech with text-to-speech, or have a chat with GPT-4o models. Pick a tool from the tabs below.

Your browser makes requests directly to OpenAI servers. We don't access or store your files, API key, or transcripts.

Authentication

Audio File

Drop audio files here or click to use file picker

Parameters

Pick the transcription model. whisper-1 is the original Whisper V2. gpt-4o-transcribe and gpt-4o-mini-transcribe support streaming and logprobs. gpt-4o-transcribe-diarize returns speaker-labeled segments.
The language of the input audio. If not specified, the model will auto-detect the language.
Optional text to guide the output style. Not supported by gpt-4o-transcribe-diarize.
Controls randomness. Higher values like 0.8 will make the output more random. Default is 0.
Output format. Whisper-1 supports json, text, srt, verbose_json, vtt. GPT-4o transcribe models only return json. The diarize model uses diarized_json.
Only available with whisper-1 in verbose_json mode.
Stream partial transcript text via server-sent events as it is generated. Ignored by whisper-1.
Return token-level log probabilities. Only available with gpt-4o-transcribe and gpt-4o-mini-transcribe in JSON format.
Controls how long audio is split. auto normalizes loudness and uses VAD. Manual lets you tune server-side VAD.

Result

API Usage Statistics

Total Requests0
Successful0
Failed0
Average Response Time0 ms
Total Audio Duration0 seconds
Estimated Cost$0.00

About this tool

This is a browser tool for testing OpenAI API endpoints. Pick a tab to switch tools:

  • Transcriptions — /v1/audio/transcriptions with whisper-1, gpt-4o-mini-transcribe, gpt-4o-transcribe, and gpt-4o-transcribe-diarize. Supports streaming, logprobs, server-VAD chunking, and known-speaker references.
  • Translations — /v1/audio/translations translates audio into English using whisper-1.
  • Text to Speech — /v1/audio/speech with tts-1, tts-1-hd, and gpt-4o-mini-tts. Returns audio you can play and download.
  • Chat Completions — /v1/chat/completions with streaming, conversation history, and token usage.

Your API key and inputs stay in your browser. Files and prompts are sent directly to OpenAI.

How It Works

  1. Get an API key — Generate one in the OpenAI dashboard.

  2. Run requests — Test OpenAI API endpoints right from your browser.

  3. Inspect, copy, and download — Per-request results, with audio playback for TTS.

Privacy & Security

Your files, prompts, and API key are never seen or processed by another party. Your browser sends data directly to OpenAI. Read OpenAI's data policy.