Category
AI speech to text APIs: what they are and who makes them
developer-first speech platforms, cloud-provider speech services, credit and bundle products, regional and domain specialists: Deepgram, Soniox, Amazon Transcribe through Scriptix.
Types of AI speech to text APIs
Developer-first speech platforms
Deepgram, AssemblyAI and Soniox sell transcription as the main product, with streaming and file endpoints and usage billing. Teams building call analytics, captions or voice interfaces usually start here, because the rate cards are built for volume and the interfaces for developers.
Cloud-provider speech services
Amazon Transcribe, Azure AI Speech and the OpenAI transcription model sit inside larger platforms. They suit teams that already hold an account there, accept the provider's billing units, and want transcription close to storage, identity and the rest of their cloud services.
Credit and bundle products
ElevenLabs draws speech to text from one shared credit pool that also pays for its other audio products, and pyannoteAI sells monthly usage credit centered on speaker diarization. Both fit buyers who want a small recurring plan before committing to volume pricing.
Regional and domain specialists
Corti targets clinical conversations, DeepScript runs on servers in Germany, and Voxist and Scriptix come from France and the Netherlands. They suit buyers whose first question is where the audio goes, or which language, industry vocabulary or compliance regime the service was built around.
Facts at a glance
| Tool | Publisher | Based in | Access |
|---|---|---|---|
| Deepgram | Deepgram (legal entity not stated on the opened pages) | Not stated on the vendor pages | Trial + API |
| Soniox | Soniox (legal entity not stated on the opened pages) | Not stated on the vendor pages | Paid + API |
| Amazon Transcribe | Amazon Web Services (legal entity not stated on the opened pages) | Not stated on the vendor pages | Trial + API |
| DeepScript | Aliru GmbH | Germany | Trial + API |
| AssemblyAI | AssemblyAI (legal entity not stated on the opened pages) | Not stated on the vendor pages | Trial + API |
| Azure AI Speech | Microsoft (legal entity not stated on the opened pages) | Not stated on the vendor pages | Free tier + API |
| OpenAI speech-to-text API (gpt-4o-transcribe) | OpenAI (legal entity not stated on the opened pages) | Not stated on the vendor pages | Paid + API |
| Corti | Corti (legal entity not stated on the opened pages) | Not stated on the vendor pages | Trial + API |
| ElevenLabs | Eleven Labs Inc. | United States | Free tier + API |
| pyannoteAI | pyannoteAI (legal entity not stated on the opened pages) | Not stated on the vendor pages | Trial + API |
| Voxist | Voxist SAS | France | Paid + API |
| Scriptix | Scriptix B.V. | Netherlands | Trial + API |
