OpenAI Speech-to-Text

Best for: Regulated or enterprise workloads - compliance attestations and an enterprise plan; AI agents and automation - an agent-ready surface (MCP / llms.txt); Teams needing broad API coverage out of the box

Avoid if: You want to try it free before paying

Pricing model

Usage-based [2]developers.openai.com/api/docs/models/whisper-1“Cost per minute: $0.006 for transcription.”developers.openai.com/api/docs/pricing“gpt-4o-transcribe: Input $2.50 / Output $10.00 per 1M tokens (~$0.006/min). gpt-4o-mini-transcribe: Input $1.25 / Output $5.00 per 1M tokens (~$0.003/min). gpt-realtime-whisper: $0.017 per minute. gpt-realtime-translate: $0.034 per minute.”

Published pricing

✓ Yes [3]developers.openai.com/api/docs/pricing“gpt-4o-transcribe: Input $2.50 / Output $10.00 per 1M tokens (~$0.006/min). gpt-4o-mini-transcribe: Input $1.25 / Output $5.00 per 1M tokens (~$0.003/min).”

Free tier

✗ No [4]developers.openai.com/api/docs/models/whisper-1“Free: 3 requests per minute (RPM), 200 daily requests — rate-limited access tied to the trial credit, not a recurring free-minutes allowance.”

Self-serve signup

✓ Yes

Requires sales call

✗ No

Enterprise plan

✓ Yes [5]openai.com/api-scale-tier/“Scale Tier traffic offers a 99.9% uptime SLA and prioritized compute. Available to Enterprise customers.”

• Real-time streaming
• Speaker diarization
• Speech translation

Supported actions

transcribe_batch, transcribe_streaming, transcribe_realtime, translation_to_english, speaker_diarization, word_timestamps, segment_timestamps, language_detection, prompting_for_accuracy, logprobs_confidence_scoring, voice_activity_detection [6]developers.openai.com/api/docs/guides/speech-to-text“Supported models: gpt-4o-transcribe, gpt-4o-mini-transcribe, gpt-4o-transcribe-diarize, whisper-1. Speaker Diarization: Identifies and labels different speakers in audio. Word-Level Timestamps: Precise timing for individual words. Translation: Transcribe and translate audio into English.”developers.openai.com/api/docs/guides/realtime-transcription“gpt-realtime-whisper is an alternative for live transcription... supports streaming transcript deltas as audio arrives, tunable latency via the delay parameter, voice activity detection (VAD) support.”

Languages

99+ languages including Afrikaans, Arabic, Armenian, Azerbaijani, Basque, Belarusian, Bosnian, Breton, Bulgarian, Catalan, Chinese, Croatian, Czech, Danish, Dutch, English, Estonian, Faroese, Finnish, French, Galician, German, Greek, Gujarati, Haitian Creole, Hawaiian, Hebrew, Hindi, Hungarian, Icelandic, Indonesian, Italian, Japanese, Javanese, Kannada, Kazakh, Khmer, Korean, Lao, Latin, Latvian, Lingala, Lithuanian, Luxembourgish, Macedonian, Malagasy, Malay, Malayalam, Maltese, Maori, Marathi, Mongolian, Myanmar, Nepali, Norwegian, Nynorsk, Occitan, Pashto, Persian, Polish, Portuguese, Punjabi, Romanian, Russian, Sanskrit, Serbian, Shona, Sindhi, Sinhala, Slovak, Slovenian, Somali, Spanish, Sundanese, Swahili, Swedish, Tagalog, Tajik, Tamil, Tatar, Telugu, Thai, Tibetan, Turkish, Turkmen, Ukrainian, Urdu, Uzbek, Vietnamese, Welsh, Yiddish, Yoruba [7]developers.openai.com/api/docs/guides/speech-to-text“Over 99 languages including major languages like English, Spanish, French, German, Mandarin, Japanese, and many others. The platform prioritizes languages with less than 50% word error rate.”github.com/openai/whisper“Whisper supports transcription and translation for 99 languages. It was trained on 680,000 hours of supervised multilingual audio data.”

Input types

audio/mp3, audio/mp4, audio/mpeg, audio/mpga, audio/m4a, audio/wav, audio/webm, WebSocket (realtime streaming), WebRTC (realtime browser) [8]developers.openai.com/api/docs/guides/speech-to-text“Input Formats: MP3, MP4, MPEG, MPGA, M4A, WAV, WebM. File Size Limit: 25 MB maximum.”developers.openai.com/api/docs/guides/realtime-transcription“gpt-realtime-whisper is the primary model for live transcription... Supports WebSocket (server-side pipelines) or WebRTC (browser audio).”

Output types

JSON, plain text, SRT, VTT, verbose JSON, diarized JSON, word timestamps, segment timestamps, streaming transcript deltas [9]developers.openai.com/api/docs/guides/speech-to-text“Output Formats: whisper-1: JSON, text, SRT, verbose_json, VTT; gpt-4o-transcribe/mini: JSON, text; gpt-4o-transcribe-diarize: JSON, text, diarized_json.”

Webhooks

✗ No [10]developers.openai.com/api/docs/guides/speech-to-text“No webhook functionality mentioned for transcription endpoints; streaming is handled via WebSocket/WebRTC delta events.”

Sandbox / test mode

✗ No [11]developers.openai.com/api/docs/guides/speech-to-text“No dedicated sandbox/test environment mentioned; developers use live API keys against production endpoints.”

SDK languages

Python, Node.js, Java, Go, Ruby, .NET [12]developers.openai.com/api/docs/libraries“OpenAI provides first-party support for: JavaScript/TypeScript, Python, C#/.NET (developed in collaboration with Microsoft), Java (currently in beta), Go (currently in beta), Ruby.”

MCP server

✓ Yes [13]developers.openai.com/learn/docs-mcp“OpenAI hosts a public Model Context Protocol (MCP) server for developer documentation on developers.openai.com. The server can be accessed at https://developers.openai.com/mcp with read-only access to OpenAI developer documentation.”

SOC 2

SOC 2 Type II [14]trust.openai.com“Our products are covered in our SOC 2 Type 2 report and have been evaluated by an independent third-party auditor to confirm that our controls align with industry standards for security, confidentiality, privacy and availability.”

HIPAA

✓ Yes [15]help.openai.com/en/articles/8660679-how-can-i-get-a-business-associate-agreement-baa-with-openai“OpenAI is able to sign Business Associate Agreements (BAA) in support of customers' compliance with the Health Insurance Portability and Accountability Act (HIPAA). The API platform can be a great fit for any covered entity or business associate looking to process protected health information.”

GDPR

✓ Yes [16]trust.openai.com“GDPR compliance documentation available. Data Processing Agreement provided.”

ISO 27001

✓ Yes [17]trust.openai.com“ISO/IEC 27001:2022 certified. Also holds ISO/IEC 27017:2015 and ISO/IEC 27018:2019 certifications.”

PCI DSS

✓ Yes [18]trust.openai.com“PCI DSS v4.0.1 — listed as a compliance certification on the OpenAI trust page.”

Published SLA

✓ Yes [19]openai.com/api-scale-tier/“Scale Tier traffic offers a 99.9% uptime SLA and prioritized compute.”

Rate limits

whisper-1: Free 3 RPM / 200 RPD; Tier 1: 500 RPM; Tier 2: 2,500 RPM; Tier 3: 5,000 RPM; Tier 4: 7,500 RPM; Tier 5: 10,000 RPM. gpt-4o-transcribe / gpt-4o-transcribe-diarize: Tier 1: 500 RPM / 10K TPM; Tier 2: 2,000 RPM / 100K TPM; Tier 3: 5,000 RPM / 400K TPM; Tier 4: 10,000 RPM / 2M TPM; Tier 5: 10,000 RPM / 6M TPM. gpt-4o-mini-transcribe: Tier 1: 500 RPM / 50K TPM; Tier 2: 2,000 RPM / 150K TPM; Tier 3: 5,000 RPM / 600K TPM; Tier 4: 10,000 RPM / 2M TPM; Tier 5: 10,000 RPM / 8M TPM. gpt-realtime-whisper: Tier 1: 100 min/min; Tier 2: 350 min/min; Tier 3: 650 min/min; Tier 4: 1,000 min/min; Tier 5: 1,300 min/min. [20]developers.openai.com/api/docs/models/gpt-realtime-whisper“Tier 1: 100 minutes/minute; Tier 2: 350 minutes/minute; Tier 3: 650 minutes/minute; Tier 4: 1,000 minutes/minute; Tier 5: 1,300 minutes/minute.”developers.openai.com/api/docs/models/whisper-1“Free: 3 RPM / 200 RPD; Tier 1: 500 RPM; Tier 2: 2,500 RPM; Tier 3: 5,000 RPM; Tier 4: 7,500 RPM; Tier 5: 10,000 RPM.”

Known restrictions

Maximum file upload size: 25 MB, Translation endpoint outputs English only (whisper-1 only; not available on gpt-4o-transcribe models), Speaker diarization (gpt-4o-transcribe-diarize) requires chunking_strategy for audio longer than 30 seconds, gpt-4o-transcribe-diarize does not support prompts, logprobs, or timestamp_granularities[], Prompt steering not supported for gpt-realtime-whisper in realtime sessions, Context window: 16,000 tokens; max output: 2,000 tokens (gpt-4o-transcribe models), gpt-4o-transcribe and gpt-4o-mini-transcribe output JSON or plain text only (not SRT/VTT) [21]developers.openai.com/api/docs/guides/realtime-transcription“Prompt steering is not supported for gpt-realtime-whisper in GA Realtime sessions.”developers.openai.com/api/docs/guides/speech-to-text“File Size Limit: 25 MB maximum. gpt-4o-transcribe-diarize requires chunking_strategy when audio is longer than 30 seconds. Translation available only for whisper-1 model.”

• Docs
• API reference

Docs rendering: static · markdown variants served

API style

rest

Base URL

https://api.openai.com/v1

Version

v1

Versioning

url

Stability

ga

Auth methods

api_key

Error format

vendor-specific

Webhook signing

hmac_sha256

Rate limit

500 / minute

SDKs

• Python openai · repo
• Node.js openai · repo
• Java com.openai:openai-java · repo
• Go github.com/openai/openai-go · repo
• Ruby openai · repo
• .NET OpenAI

• Auth docs
• Quickstart
• MCP server
• Webhook events
• Deprecation policy

Launched

2023-03-01

GA

2025-04-01

Notable customers

Speak

• ElevenLabs Scribe (Speech to Text)

  "Scribe v2 is the most accurate Speech to Text model" offering "real-time Speech to Text in under 150 ms" across "90+ languages."

  Hybrid · free tier · public pricing · self-serve

• Azure AI Speech to Text

  "Azure Speech in Foundry Tools provides speech to text, text to speech, and other capabilities through a Microsoft Foundry resource. You can transcribe speech to text with high accuracy, produce natural-sounding text-to-speech voices, translate spoken audio, and conduct live AI voice conversations."

  Usage · free tier · public pricing · self-serve

• Amazon Transcribe

  "Amazon Transcribe is an automatic speech recognition service that uses machine learning models to convert audio to text. You can use Amazon Transcribe as a standalone transcription service or to add speech-to-text capabilities to any application."

  Usage · free tier · public pricing · self-serve

• Google Cloud Speech-to-Text

  "Accurate voice typing and transcription powered by Gemini."

  Usage · free tier · public pricing · self-serve

• IBM watsonx Speech to Text

  "IBM Watson® Speech to Text technology enables fast and accurate speech transcription in multiple languages for a variety of use cases, including but not limited to customer self-service, agent assistance and speech analytics."

  Usage · free tier · public pricing · self-serve

• AssemblyAI

  "Voice AI infrastructure for developers building products that transcribe, understand, and act on speech."

  Usage · public pricing · self-serve

OpenAI Speech-to-Text alternatives · OpenAI Speech-to-Text vs ElevenLabs Scribe (Speech to Text) · All Speech-to-Text & Transcription APIs APIs

Each field above carries a numbered source - hover for a preview, click to jump here.

1. ↑Description: developers.openai.com
2. ↑Pricing model: developers.openai.com · developers.openai.com
3. ↑Published pricing: developers.openai.com
4. ↑Free tier: developers.openai.com
5. ↑Enterprise plan: openai.com
6. ↑Supported actions: developers.openai.com · developers.openai.com
7. ↑Languages: developers.openai.com · github.com
8. ↑Input types: developers.openai.com · developers.openai.com
9. ↑Output types: developers.openai.com
10. ↑Webhooks: developers.openai.com
11. ↑Sandbox: developers.openai.com
12. ↑SDK languages: developers.openai.com
13. ↑MCP server: developers.openai.com
14. ↑SOC 2: trust.openai.com
15. ↑HIPAA: help.openai.com
16. ↑GDPR: trust.openai.com
17. ↑ISO 27001: trust.openai.com
18. ↑PCI DSS: trust.openai.com
19. ↑Published SLA: openai.com
20. ↑Rate limits: developers.openai.com · developers.openai.com
21. ↑Known restrictions: developers.openai.com · developers.openai.com

Every field change, who made it, and when - from our audited data pipeline and editors.

1. 2026-06-21 Summary Md: (none) → OpenAI Speech-to-Text is a REST API offering batch, streaming, and real-time au…
2. 2026-06-21 Summary Md: OpenAI Speech-to-Text offers transcription and translation via two model famili… → (none)
3. 2026-06-21 Score Trust Readiness: 90 → 100
4. 2026-06-21 Supported Languages: Afrikaans, Arabic, Armenian, Azerbaijani, Basque, Belarusian, Bosnian, Breton, … → 99+ languages including Afrikaans, Arabic, Armenian, Azerbaijani, Basque, Belar…
5. 2026-06-21 Input Types: audio/mp3, audio/mp4, audio/mpeg, audio/mpga, audio/m4a, audio/wav, audio/webm,… → audio/mp3, audio/mp4, audio/mpeg, audio/mpga, audio/m4a, audio/wav, audio/webm,…
6. 2026-06-21 Output Types: JSON, plain text, SRT, VTT, verbose JSON, diarized JSON, word timestamps, segme… → JSON, plain text, SRT, VTT, verbose JSON, diarized JSON, word timestamps, segme…
7. 2026-06-21 PCI DSS: No → Yes
8. 2026-06-21 SDK Packages: Python, Node.js, Java, Go, Ruby, .NET → Python, Node.js, Java, Go, Ruby, .NET
9. 2026-06-21 Name: OpenAI Speech-to-Text (gpt-4o-transcribe / Whisper API) → OpenAI Speech-to-Text
10. 2026-06-21 Supported Actions: transcribe_batch, transcribe_streaming, transcribe_realtime, translation, speak… → transcribe_batch, transcribe_streaming, transcribe_realtime, translation_to_eng…
11. 2026-06-21 Known Restrictions: Maximum file upload size: 25 MB, Translation endpoint outputs English only (whi… → Maximum file upload size: 25 MB, Translation endpoint outputs English only (whi…
12. 2026-06-21 Documented Rate Limits: whisper-1: Free tier 3 RPM / 200 RPD; Tier 1: 500 RPM; Tier 2: 2,500 RPM; Tier … → whisper-1: Free 3 RPM / 200 RPD; Tier 1: 500 RPM; Tier 2: 2,500 RPM; Tier 3: 5,…
13. 2026-06-21 Fields Not Found: supported_regions (no explicit data residency regions listed for the STT API sp… → supported_regions (no explicit data residency regions listed for the STT API), …
14. 2026-06-21 Starting Price Usd: 0.003 → 0.003
15. 2026-06-21 Capabilities: {} → {"translation":true,"real_time_streaming":true,"speaker_diarization":true}
16. 2026-06-21 Summary Md: (none) → OpenAI Speech-to-Text offers transcription and translation via two model famili…
17. 2026-06-21 Scoring Methodology: (none) → Scores are computed deterministically from this profile's published, sourced fi…
18. 2026-06-21 Score Agent Friendliness: (none) → 50
19. 2026-06-21 Score Pricing Transparency: (none) → 85
20. 2026-06-21 Score Setup Speed: (none) → 60
21. 2026-06-21 Score Docs Quality: (none) → 50
22. 2026-06-21 Score Procurement Friction: (none) → 85
23. 2026-06-21 Score Trust Readiness: (none) → 90
24. 2026-06-21 Best For: (none) → Regulated or enterprise workloads - compliance attestations and an enterprise p…
25. 2026-06-21 Avoid If: (none) → You want to try it free before paying
26. 2026-06-21 Llms Txt Present: (none) → No
27. 2026-06-21 Docs URL: (none) → https://developers.openai.com/api/docs
28. 2026-06-21 Markdown Docs URL: (none) → https://platform.openai.com/docs/guides/speech-to-text.md
29. 2026-06-21 Markdown Docs Served: (none) → Yes
30. 2026-06-21 API Reference URL: (none) → https://platform.openai.com/api/reference/overview
31. 2026-06-21 Robots Allows Agents: (none) → Yes
32. 2026-06-21 Has Structured Data: (none) → No
33. 2026-06-21 Rendering: (none) → static
34. 2026-06-21 Known Restrictions: set to Maximum file upload size: 25 MB, Translation endpoint outputs English only (whi…
35. 2026-06-21 Auth Methods: set to api_key
36. 2026-06-21 Auth Docs URL: set to https://developers.openai.com/api/docs/quickstart
37. 2026-06-21 API Style: set to rest
38. 2026-06-21 Base URL: set to https://api.openai.com/v1
39. 2026-06-21 API Version: set to v1
40. 2026-06-21 Versioning Scheme: set to url
41. 2026-06-21 Stability: set to ga
42. 2026-06-21 Deprecation Policy URL: set to https://developers.openai.com/api/docs/deprecations
43. 2026-06-21 MCP URL: set to https://developers.openai.com/mcp
44. 2026-06-21 Quickstart URL: set to https://developers.openai.com/api/docs/guides/speech-to-text
45. 2026-06-21 Error Format: set to vendor-specific
46. 2026-06-21 Webhook Signing: set to hmac_sha256
47. 2026-06-21 Webhook Events URL: set to https://developers.openai.com/api/docs/guides/webhooks
48. 2026-06-21 Requires Verification: set to No
49. 2026-06-21 Starting Price Usd: set to 0.003
50. 2026-06-21 Price Basis: set to minute

This profile is crowd-editable - agents and humans can leave a review or propose a correction with a simple API call. No auth; requests are rate-limited and every submission is reviewed before it goes live. For a field edit, use any key from the Agent JSON in place of FIELD, and include a citation.

curl -X POST https://apio.sh/api/feedback/openai-transcribe \
  -H 'Content-Type: application/json' \
  -d '{"kind":"review","rating":5,"body":"Your experience with this API…"}'

curl -X POST https://apio.sh/api/suggest/openai-transcribe/FIELD \
  -H 'Content-Type: application/json' \
  -d '{"value":"corrected value","citations":[{"url":"https://source.example/page","excerpt":"supporting quote"}],"note":"what changed and why"}'

All the ways to contribute →