Use the OpenAI Transcription API via RunAPI with a model skill, unified auth, and pay-as-you-go pricing.
curl -X POST https://runapi.ai/v1/audio/transcriptions \
-H "Authorization: Bearer $RUNAPI_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "whisper-1",
"audio_url": "https://cdn.runapi.ai/public/samples/voice.mp3"
}'import { OpenaiTranscriptionClient } from "@runapi.ai/openai-transcription";
const client = new OpenaiTranscriptionClient();
const result = await client.speechToText.run({
model: "whisper-1",
audio_url: "https://cdn.runapi.ai/public/samples/voice.mp3",
});require "runapi/openai_transcription"
client = RunApi::OpenaiTranscription::Client.new
result = client.speech_to_text.run(
model: "whisper-1",
audio_url: "https://cdn.runapi.ai/public/samples/voice.mp3"
)npx skills add runapi-ai/openai-transcription -g# Claude Code
claude mcp add runapi -s user -- npx -y @runapi.ai/mcp
# Codex
codex plugin install runapi-mcp@agents
# Cursor / Windsurf / VS Code
npx @runapi.ai/mcp init cursorOpenAI Transcription converts uploaded audio files into text through a synchronous multipart endpoint. Whisper-1 supports subtitle and timestamp output, while GPT Transcribe supports keyword and language hints for multilingual audio.
| Endpoint | Resolution | Duration | Price | |
|---|---|---|---|---|
| speech_to_text | — | — | $0.02 / minute |
Prices shown for gpt-transcribe. Other variants retain their own endpoint pricing.
Sign up and generate an API key from the dashboard.
Choose the version that best matches your quality and cost requirements.
Submit your prompt and parameters to the model's documented endpoint.
Poll the task or set a webhook, then download the completed output.
OpenAI Transcription provides speech-to-text models for meeting notes, captions, searchable media, and multilingual transcripts. Both variants use the OpenAI-compatible audio transcription request shape.
A single RunAPI API key unlocks the whole model catalog across all providers. No separate accounts to create, no API keys to rotate per integration, and no credential management overhead. Add a new model to your app by changing one parameter.
Per-call pricing in USD, billed monthly into a single invoice. No subscription tiers, no minimum spend, and failed generations are never charged. The pricing page and check_pricing API show exact costs before you commit to a model.
Typed schemas, parameter constraints, and setup notes are packaged in the model skill so your implementation starts from the right contract. The skill loads into Claude Code, Codex, Gemini CLI, Cursor, and VS Code — your agent knows the correct request shape before you write a line of code.
Whisper-1 supports JSON, plain text, subtitle, verbose JSON, and timestamped output. GPT Transcribe focuses on JSON or plain text and accepts keyword and multiple-language hints.
The endpoint accepts FLAC, MP3, MP4, MPEG, MPGA, M4A, OGG, WAV, and WebM uploads up to 25 MB.
Whisper-1 can return SRT or VTT subtitles and word- or segment-level timestamps through verbose JSON. Those output modes are not available with GPT Transcribe.
Yes. You can provide one language hint or a list of likely languages, plus keyword hints for names and domain-specific terms.
Yes. A successful request returns the completed transcript directly, without a separate task polling endpoint.
Pick the cheapest variant that meets your quality bar. Most teams start on the fast variant and graduate to pro for production.