Hindi speech-to-text APIs compared: prices and features, October 2026.

What does each Hindi speech-to-text API cost?

List prices, read on 16 October 2026 from the linked page, in each vendor’s currency, without volume deals. A dash means the pages didn’t say.

Provider and model	Published price	Hindi, as published	Live streaming	Source
Frequency, Fourier Transcribe	₹18 per hour of recorded audio	Built for Hindi and Hinglish phone calls; also Gujarati, Tamil and Telugu; detects the main language	Standalone streaming in private preview; live inside every agent	Docs
Sarvam, Saaras	₹0.49 a minute real-time or batch, pay as you go; ₹0.41 on Growth and Business; batch with diarization ₹0.59	23 languages, 22 Indian plus English; a codemix output mode	WebSocket streaming	Pricing, models
ElevenLabs, Scribe v2; Scribe v2 Realtime	$0.22; $0.39 per hour	Hindi among 90+ languages	Scribe v2 Realtime	Pricing, languages
Deepgram, Nova-3	Pre-recorded $0.0043 a minute (multilingual $0.0052); streaming $0.0048 (limited-time; regular $0.0077)	Hindi (hi); the multilingual mode covers Hindi with English	WebSocket streaming	Pricing, languages
Google Cloud, Speech-to-Text V2	Standard $0.016 a minute for the first 500,000 minutes a month; dynamic batch $0.003	hi-IN on chirp_3 (us, eu regions), chirp_2 and telephony models	—	Pricing, languages
Microsoft Azure, Speech	Real-time $1 per hour; fast $0.36; batch $0.18	hi-IN listed, including fast transcription	Real-time transcription	Pricing, languages
Amazon Transcribe	US East examples: streaming $0.01 a minute, batch $0.006; tiered and varies by region	Hindi (hi-IN), batch and streaming	Streaming	Pricing, languages
OpenAI, gpt-transcribe; gpt-4o-transcribe; gpt-4o-mini-transcribe	$0.0045; $0.006; $0.003 a minute (estimated)	Language hints by ISO code; whisper-1 follows Whisper’s 98 languages	gpt-live-transcribe, $0.017 a minute	Pricing, guide

Speaker labels are included in our ₹18. Sarvam prices batch with diarization separately, Deepgram and Azure charge an add-on, and OpenAI uses a separate model, gpt-4o-transcribe-diarize.

What does an hour of Hindi audio cost?

At list price, before tax and free tiers:

Provider and model	Working	Cost
Fourier Transcribe	1 × ₹18	₹18
Sarvam Saaras, batch, pay as you go	60 × ₹0.49	₹29.40
ElevenLabs Scribe v2	1 × $0.22	$0.22
Deepgram Nova-3 Monolingual, pre-recorded	60 × $0.0043	$0.258
Google Speech-to-Text V2, dynamic batch	60 × $0.003	$0.18
Azure batch transcription	1 × $0.18	$0.18
Amazon Transcribe batch, US East example	60 × $0.006	$0.36
OpenAI gpt-transcribe	60 × $0.0045	$0.27

Convert the dollar figures at the rate you pay. At small volume, free allowances matter: Azure gives 5 hours of real-time transcription a month, Amazon 60 minutes a month for a year, and Deepgram $200 of credit.

Which one is most accurate?

Vendors measure accuracy their own way, so we don’t repeat their figures. Ours comes from a report on Fourier STT, our own Hindi speech recognition: all 10,000 utterances of the public RinggAI Hindi benchmark, six corpora, five systems, the same references and the same normaliser.

System	Normalised WER, all 10,000 utterances
Fourier STT	9.12%
Ringg Parrot	9.72%
ElevenLabs	11.11%
Sarvam	12.83%
Deepgram	15.79%

Read the limits with the numbers. The vendor scores are outputs Ringg bundled with the dataset, and may not match what each vendor serves today. We weren’t lowest everywhere: Ringg scored lower on CommonVoice and IndicTTS, and ElevenLabs on FLEURS. A second build, the one in our package, scores 9.85%. On Lahaja, a test-only corpus, Fourier STT scored 11.63% against 16.55% for the open Qwen3-ASR-1.7B, the only other system run there.

This is all read and prompted speech, so it says nothing about an 8 kHz phone call. Fourier Transcribe, our product for recordings, adds more passes and a language model on top, so the table describes the recogniser alone. The method is on the Fourier STT page.

How do I test it on my own audio?

1. Pick twenty to fifty recordings that sound like production.
2. Have a Hindi speaker write reference transcripts. Agree first how English words are written: “ऑर्डर” and “order” are one word to a listener and two to a WER script.
3. Normalise spellings that sound the same, like अन्दर and अंदर. Leave grammar alone: है and हैं stay different words.
4. Send the same files and language setting to each service, plus any vocabulary it accepts.
5. Score word error rate, then read the worst ten yourself, watching names, numbers and amounts.
6. Check speaker labels separately. Agent and customer swapped is still wrong.

With us, upload each file to POST /v2/transcriptions with a line of context and your vocabulary, then read the job until it completes; the transcription guide has the requests.

Building a calling agent? Compare per-minute plans instead. Ours cost ₹1 a minute on the volume plan, recognition included.