AWS Marketplace now accepts line of credit payments through the PNC Vendor Finance program. This program is available to select AWS customers in the US, excluding NV, NC, ND, TN, & VT.

View financing details

Pricing

Free trial

Try for free

Try this product free for 7 days according to the free trial terms set by the vendor.

Kotoba Speech-to-Text (STT) Streaming

Info

View purchase options

Pricing is based on actual usage, with charges varying according to how much you consume. Subscriptions have no end date and may be canceled any time.

Additional AWS infrastructure costs may apply. Use the AWS Pricing Calculator to estimate your infrastructure costs.

Usage costs (11)

Info

Dimension	Description	Cost/host/hour
ml.c4.2xlarge Inference (Batch) Recommended	Model inference on the ml.c4.2xlarge instance type, batch mode	$0.00
ml.g6e.8xlarge Inference (Real-Time) Recommended	Model inference on the ml.g6e.8xlarge instance type, real-time mode	$38.40
ml.g6e.xlarge Inference (Real-Time)	Model inference on the ml.g6e.xlarge instance type, real-time mode	$38.40
ml.g6e.2xlarge Inference (Real-Time)	Model inference on the ml.g6e.2xlarge instance type, real-time mode	$38.40
ml.g6e.4xlarge Inference (Real-Time)	Model inference on the ml.g6e.4xlarge instance type, real-time mode	$38.40
ml.g6e.16xlarge Inference (Real-Time)	Model inference on the ml.g6e.16xlarge instance type, real-time mode	$38.40
ml.g6.xlarge Inference (Real-Time)	Model inference on the ml.g6.xlarge instance type, real-time mode	$4.80
ml.g6.2xlarge Inference (Real-Time)	Model inference on the ml.g6.2xlarge instance type, real-time mode	$4.80
ml.g6.4xlarge Inference (Real-Time)	Model inference on the ml.g6.4xlarge instance type, real-time mode	$4.80
ml.g6.8xlarge Inference (Real-Time)	Model inference on the ml.g6.8xlarge instance type, real-time mode	$4.80

AI Insights

Info

Dimensions summary

You pay by the hour for each active inference instance (HostHrs), so cost tracks how long you run the model. The dimensions split into two modes. Batch mode runs on the ml.c4.2xlarge CPU instance for processing groups of audio. Real-time mode runs on GPU instances for live speech-to-text. Real-time offers two GPU families, each in sizes from xlarge up to 16xlarge. Larger instance sizes add compute capacity and bill at higher hourly rates. You choose the instance that fits your workload and scale by selecting a bigger size or running more instances.

Top-of-mind questions for buyers

What does one HostHrs unit cover, and how is it counted?

One HostHrs equals one hour of a single running inference instance of the chosen type. You are billed for each hour that instance is active. Running two instances of the same type doubles the hourly count. Time accrues while the instance runs, regardless of how much audio you process.

Am I charged when an inference instance is stopped or idle?

Software charges accrue per active instance-hour. A fully stopped instance stops accruing software charges. Underlying AWS infrastructure fees may still apply for stopped-but-allocated resources, but the listing meters running instance time only. To stop billing, shut the instance down.

Should I pick a batch or real-time instance, and how do they differ mechanically?

The ml.c4.2xlarge batch instance runs on CPU for processing groups of recorded audio. The real-time GPU instances handle live streaming speech-to-text. Batch suits offline transcription jobs. Real-time suits interactive voice applications needing low latency. You choose based on whether audio is pre-recorded or streamed live.

site.kotoba.tech

Helpful?

Vendor refund policy

n/a

How can we make this page better?

Tell us how we can improve this page, or report an issue with this product.

Legal

Vendor terms and conditions

Upon subscribing to this product, you must acknowledge and agree to the terms and conditions outlined in the vendor's End User License Agreement (EULA) .

Content disclaimer

Vendors are responsible for their product descriptions and other product content. AWS does not warrant that vendors' product descriptions or other product content are accurate, complete, reliable, current, or error-free.

Usage information

Info

Version

Delivery details

Amazon SageMaker model

An Amazon SageMaker model package is a pre-trained machine learning model ready to use without additional training. Use the model package to create a model on Amazon SageMaker for real-time inference or batch processing. Amazon SageMaker is a fully managed platform for building, training, and deploying machine learning models at scale.

Deploy the model on Amazon SageMaker AI using the following options:

Real-time inference

Deploy the model as an API endpoint for your applications. When you send data to the endpoint, SageMaker processes it and returns results by API response. The endpoint runs continuously until you delete it. You're billed for software and SageMaker infrastructure costs while the endpoint runs. AWS Marketplace models don't support Amazon SageMaker Asynchronous Inference. For more information, see Deploy models for real-time inference .

Batch transform

Deploy the model to process batches of data stored in Amazon Simple Storage Service (Amazon S3). SageMaker runs the job, processes your data, and returns results to Amazon S3. When complete, SageMaker stops the model. You're billed for software and SageMaker infrastructure costs only during the batch job. Duration depends on your model, instance type, and dataset size. AWS Marketplace models don't support Amazon SageMaker Asynchronous Inference. For more information, see Batch transform for inference with Amazon SageMaker AI .

Version release notes

Improve inference stability in long-running sessions (over 120 seconds)
Remove 300 seconds of session limit
Improve Japanese transcription accuracy and noise robustness
Add more sample rates and audio formats
Transcribe person's names in Kata-kana
Add keywords support
Add server-side turn detection (server VAD)

Additional details

Inputs

Summary: Model input summary

This model accepts audio data via SageMaker bidirectional streaming using JSON events over HTTP/2.

Input Event Types:

transcription_session.update - Configure the session before sending audio

input_audio_format: Audio encoding format ("pcm16", "float32", "mulaw", or "opus")

input_audio_sample_rate: Sample rate in Hz (default: 24000)

input_audio_number_of_channels: Must be 1 (mono)

input_audio_transcription.language: Source language (ISO-639-1: "ja" or "en")

input_audio_transcription.target_language: Output language (ISO-639-1: "ja" or "en")

input_audio_transcription.kana: Transcribe person's name in Kata-kana (default: false)

input_audio_transcription.keywords: keyword list (default: null)

turn_detection: false (default) or a server VAD config object ({"type": "server_vad", "silence_duration_ms": 400}) to enable automatic turn detection

turn_detection.silence_duration_ms: Trailing silence (ms) that ends a turn (default: 400; tune per application/pipeline)

input_audio_buffer.append - Send audio chunks

audio: Base64-encoded audio bytes

event_id: Optional correlation ID for tracking (max 36 characters)

input_audio_buffer.commit - Signal end of audio stream

Supported Audio Formats:

pcm16: 16-bit signed integer PCM, little-endian, mono

float32: 32-bit floating point, little-endian, mono, range [-1.0, 1.0]

mulaw: G.711 μ-law

opus: Self-contained Ogg/Opus blob per append event

Recommended Settings:

Sample rate: 24000 Hz

Chunk duration: 80ms per event

Limitations for input type: - Maximum audio per event: 8 seconds (~1MB base64 encoded) - Supported language: ja

Input MIME type: application/json

Real-time inference sample input data

https://github.com/kotoba-tech/api-samples/blob/main/stt/data/sample_input.json

Batch transform sample input data

Not support

Support

Vendor support

For support, contact: kotoba_product+marketplace@kotoba.tech

AWS infrastructure support

AWS Support is a one-on-one, fast-response support channel that is staffed 24x7x365 with experienced and technical support engineers. The service helps customers of all sizes and technical abilities to successfully utilize the products and features provided by Amazon Web Services.

Get support

Similar products

Kotoba Text-to-Speech (TTS) Streaming

By Kotoba Technologies, Inc.

Real-time Japanese text-to-speech via SageMaker bidirectional streaming. Sub-second first-audio latency.

View product

Customer reviews

Leave a review

Ratings and reviews

Info

0 ratings

5 star

4 star

3 star

2 star

1 star

0 reviews

No customer reviews yet

Be the first to review this product . We've partnered with PeerSpot to gather customer feedback. You can share your experience by writing or recording a review, or scheduling a call with a PeerSpot analyst.