Against models that hear the whole recording
Word error rate, average of 10 Indian languages, lower is better. Every system that takes a language setting was given it.
9%fewer errors than the best batch model, while streaming
Chariot-1-ASR Realtime
Audio and video to text, in the language it was spoken.
Upload a recording, or a video with sound, and get its transcript back in Hindi, English or any of 12 languages, with English words kept as they were said. Copy it, download it, or download subtitles from its word timings.
Free to start · No card · Audio or video files · 12 languages
0:00
The clip
English words
Clips: FLEURS, CC BY 4.0
#1in all 10 Indian languages we tested, against every model that hears the whole recording
12languages, detected from the audio when you leave the language out
MP4and other video files read directly, as long as they have an audio track
3 ha recording through the async API, up to 100 recordings in a batch
Benchmark · October 2026
#1 in all 10 Indian languages. Against the batch models too.
Files run in high-accuracy mode. Here is the same model against the models built for whole recordings.
How it works
Upload, then read.
Upload the file
In Speech to Text in the web app, drop in a recording or a video: MP3, WAV, M4A, MP4 and more.
Choose the settings
Set the language or leave it out, choose codemix or native for English words, and turn on word timestamps for subtitles.
Copy or download
Copy the transcript, download it as text, or download an SRT file from its word timings. Every transcript is saved in your history.
API
From a file to its text.
One request a file: send the recording, or a link to it, and the complete transcript comes back. Add include_timestamps for each word’s time.
curl https://api.chariot.in/v1/asr/transcribe \
-H "chariotai-api-key: $CHARIOT_API_KEY" \
-F "language=hi" \
-F "include_timestamps=true" \
-F "input_audio_file=@recording.mp4"
# {"text": "...", "duration": 42.6,
# "words": [{"text": "...", "start": 0.38, "end": 0.71}, ...]}Recordings
Short clips or three-hour recordings.
The web app for a file at a time, the API for everything else.
In the web app
Upload a file and read its transcript in your browser. Transcripts stay in your history.
One request a file
The file API takes a recording, or a link to it, and returns the complete transcript in one HTTP response.
Batches
The async API takes up to 100 recordings in one batch, each up to three hours and 1 GiB, and calls a webhook when they finish.
Audio and video
WAV, MP3, FLAC, OGG, M4A, AAC and WebM, and video such as MP4 with an audio track.
Word timings
Ask for timestamps and each word comes back with its start and end time, estimated from the audio.
English words
Codemix keeps borrowed English words in Latin letters; native writes them in the script of the spoken language.
Pricing
Plans from ₹499 a month.Plans from $5 a month. Or start free.
Speech to text uses the same credits as text to speech, at one rate in both modes. Each plan covers this much audio a month.
| Plan | A month | Speech to text |
|---|---|---|
| Free10,000 credits | ₹0$0 | about 2 hours |
| Starter100,000 credits | ₹499$5 | about 19 hours |
| Creator250,000 credits | ₹999 | about 46 hours |
| Startup1,500,000 credits | ₹4,999$49 | about 280 hours |
| Scale10,000,000 credits | ₹29,999$299 | about 1,900 hours |
Monthly prices. Prices exclude applicable taxes.
Use cases
Other things people build with it.
Voice agents, too: they hear a caller while they are still speaking, in their language.
Call transcription
Support and sales calls written down, live from the phone line or from recordings.
Call transcription ›Voice typing
Speak into the microphone, watch the words appear, and copy them anywhere.
Voice typing ›Your data
Your audio stays in India.
Speech to text runs on Indian infrastructure. Audio you send and its transcripts are processed in India under the Digital Personal Data Protection Act. Chariot is one of four teams selected under the IndiaAI Mission to build the country’s foundation models.
IndiaAudio and transcripts are processed in India.
DPDPHandled under India’s data protection law.
1 of 4Teams selected under the IndiaAI Mission to build foundation models.
Can I transcribe a video?
Yes. Upload the video, such as an MP4, or its audio. Chariot reads the audio track, so there is nothing to convert first.
Which file formats work?
WAV, MP3, FLAC, OGG, M4A, AAC and WebM, and video containers such as MP4 that have an audio track.
How long can a recording be?
Through the async API, up to three hours and 1 GiB a recording, and up to 100 recordings in a batch. A live WebSocket session also runs for up to three hours.
Does it translate?
No. It writes down what was said, in the language it was said. Setting a language tells the model what to expect; it does not translate.
Does it label who is speaking?
Not yet. Transcripts are plain text, with optional word timestamps.
How is it billed?
1.5 credits per second of audio, 90 credits a minute, from the same credits as text to speech. Every account starts with 10,000 free credits, almost two hours of audio.
Where is my audio processed?
In India. Audio you send and its transcripts are processed on Indian infrastructure, under the Digital Personal Data Protection Act.
Which languages does it support?
Hindi, Marathi, Bengali, Tamil, Telugu, Kannada, Malayalam, Gujarati, Punjabi, Odia, Sanskrit and English. Set the language for a session, or leave it out and the model detects it from the audio.
