AI Speech to Text Tools
PRODUCT OVERVIEW
OCI Speech is an Oracle Cloud Infrastructure AI service that transcribes speech to text and synthesizes natural-sounding speech from text. It provides accurate, text-normalized, time-stamped transcriptions and synthetic voice through the OCI Console, OCI Data Science notebooks, REST APIs, CLIs, and SDKs. Prebuilt acoustic and language models require no data science experience, and developers can integrate the service without ML expertise. Features include asynchronous file-based transcription, limited-availability real-time transcription, neural text-to-speech, diarization, word-level confidence scores, profanity filters, and multilingual support for English, Spanish, Portuguese, German, French, Italian, and Hindi, plus the OpenAI Whisper model for 57+ languages. Typical uses include closed captioning, call analytics, medical dictation, and accessibility. Billing is per transcription hour with five free hours per month per tenancy and no setup charges or minimum commitments.
Use the Oracle Cloud sign-in page to access the OCI Console and begin using OCI Speech.
Submit single-channel 16-bit PCM WAV files at 16 kHz, or use supported formats such as MP3, MP4, FLAC, and others that are converted to PCM WAV before transcription.
Start asynchronous file-based transcription through the OCI Console, REST APIs, CLIs, SDKs, or an OCI Data Science notebook session.
Receive JSON by default or request SRT output at no additional cost.
Use the limited-availability neural text-to-speech feature to generate human-like speech and select from available voices.
Digital media content searching and closed captions
Customer feedback and call analytics
Medical dictation
Improved accessibility with neural text-to-speech
Indexing and analyzing audio and video files
OCI Speech is an Oracle Cloud Infrastructure AI service that transcribes speech to text and synthesizes speech from text using automatic speech recognition and neural text-to-speech.
Precision billing charges US$0.50 per hour of transcription or voice synthesis, measured in seconds. Five free transcription hours are included per month per tenancy.
No. OCI Speech has no setup charges or minimum service commitments and requires no hardware.
Pretrained models support English, Spanish, Portuguese, German, French, Italian, and Hindi. The OpenAI Whisper model supports 57+ languages for asynchronous file-based transcription.
Single-channel 16-bit PCM WAV at 16 kHz is supported directly. AAC, AC3, AMR, AU, FLAC, M4A, MKV, MP3, MP4, OGA, OGG, WAV, and WEBM files are converted to PCM WAV before transcription.
JSON is the default output, and SRT is available as an option at no additional cost.
No. OCI Speech only transcribes your content and keeps no information from the file for training, debugging, or other purposes.
Real-time transcription is offered in limited availability. File-based asynchronous transcription is currently supported.
Billing uses precision billing at US$0.50 per transcription hour, measured in seconds. Five free transcription hours are included per month per tenancy. No setup charges or minimum service commitments apply. Text-to-speech and real-time transcription are in limited availability.
Five hours of free transcription every month per tenancy, with no setup charges or minimum commitments.
Precision billing charges US$0.50 per hour of transcription or voice synthesis, measured in seconds after the free monthly hours.
COMMUNITY REVIEWS
No reviews yet.
Speech-to-text API for real-time and pre-recorded transcription in 55+ languages, with medical and voice-agent models and secure deployment options.
Azure Speech in Foundry Tools provides APIs and SDKs for speech-to-text, text-to-speech, speech translation, and speaker recognition.
Google Cloud Speech-to-Text converts audio into text with APIs, streaming recognition, 85+ languages, and speaker diarization.
