If you're looking for a speech-to-text tool with good pricing options, AssemblyAI is a strong contender. AssemblyAI offers a range of AI models for speech-to-text transcription, speaker identification and sentiment analysis, trained on 12.5 million hours of multilingual audio data. The service supports more than 99 languages and offers integration tools, including a free tier for prototyping and pay-as-you-go pricing that starts at $0.12 per hour. Discounts for large volumes also are available, so it's a good option for many use cases.
Another strong contender is SpeechText, which transcribes audio and video into text with high accuracy. It supports more than 30 languages and offers domain-specific models for better performance in areas like journalism and medicine. SpeechText offers STARTER, PERSONAL, STANDARD and BUSINESS pricing tiers for different needs and budgets. The service offers an API for integration with your own apps, and it's GDPR compliant to protect your data.
Rev AI strikes a balance between high accuracy and flexibility with its speech-to-text API. It offers asynchronous, streaming and human transcription options in multiple languages. Pricing is pay-as-you-go, with machine transcription costing $0.02 per minute and human transcription costing $1.50 per minute. That's good for a variety of industries, including media, education and call centers, where accessibility and efficiency are important.
If you need low-latency and low-cost options, Deepgram is worth a look. Deepgram's suite of APIs includes speech-to-text, text-to-speech and audio intelligence, and it supports multiple languages with high accuracy. It offers a free $200 credit to get you started and flexible pricing tiers to accommodate your needs. It's good for speech analytics and media transcription, and it's got low latency and integration options.