Sarvam AI 发布语音识别模型 Saaras V4

The Next Leap in ASR for a Multilingual World

AnnouncementProduct

September 25, 2026 · 6 min read

Introducing Saaras V4

We are releasing Saaras V4, the latest generation of our speech recognition model.

This post covers the work behind Saaras V4, including the data used to train it, our pre-training approach, the architectural and decoding changes that support multiple output modes, and the evaluation methodology behind our results. We also report benchmark performance across Indian languages, English, and challenging real-world speech conditions.

Highlights

Saaras V4 uses an audio encoder, and an LLM decoder. The LLM decoder is a 3B hybrid state space model completely trained from scratch in-house
Saaras V4 achieved SOTA performance on all 22 Indian languages and reaffirms our commitment to even the most low-resource Indian languages
Saaras V4 achieves lowest WER on seven global English datasets, six being foreign English dialects and one being Indian English
You can choose the format that works best for you: Saaras V4 supports five transcript formats: transcribe, translate, verbatim, translit, and codemix
Saaras V4 is robust to noisy audios and code mixing. It also handles dialectical variation seamlessly
Its language identification is also best-in-class: a 2.9% error rate across the top ten Indian languages, and 5.22% across all 22

Inside Saaras V4

Saaras V4 pairs an audio encoder with a 3B hybrid state-space language model that we trained from scratch in-house.

We designed the system to handle the range of speech we see in practice, including code-mixing, dialectal variation, and noisy audio, while supporting five different ways of representing the same speech: verbatim transcription, normalized text, code-mixed text, transliteration, and translation.

Figure 1. Overview of our approach: The input audio waveform goes into an audio encoder, which produces audio embeddings that capture phonetic and acoustic detail. A temporal-downsampling adapter compresses these along the time axis and projects them into the language model's embedding space, so even long recordings stay within the context budget. The in-house Sarvam-3B autoregressive decoder then reads these audio features alongside the prompt as text features, and emits the transcript token by token, with each token fed back as input for the next.

Performance across English benchmarks

We evaluated Saaras V4 across seven English speech recognition benchmarks covering Indian English, international accents, meetings, media, finance, and other real-world speech settings. Across these datasets, Saaras V4 achieves the lowest average word error rate.

AMI

Meeting rooms, far-field microphones

GigaSpeech

Podcast, audiobook, web video

LibriSpeech clean

Read speech, studio conditions

LibriSpeech others

Read speech, harder acoustics

SPGISpeech

Earning calls, financial vocabulary

VoxPopuli

European Parliament speeches

Svarah

Indian-accented English

Evaluation methodology

Six international: taken as published at HuggingFace's Open ASR Leaderboard
One Indian accented English: Svarah published by AI4Bharat
Normalization and scoring follow the Open ASR Leaderboard's code

Indic Benchmarks for Saaras V4:

Vistaar

We evaluate Saaras V4 on Vistaar across ten Indian languages and report both standard Word Error Rate (WER) and LLM-WER.

WER remains the standard deterministic measure for speech recognition, counting substitutions, deletions, and insertions between the reference and predicted transcript. In Indic-language settings, however, some of these differences can be orthographic or formatting variations that do not materially change the spoken content.

To account for this, we also report LLM-WER, which adds a semantic adjudication step to determine whether a mismatch changes the meaning of the transcript or simply represents the same content differently.

The two metrics are therefore complementary. WER provides a reproducible baseline, while LLM-WER gives a more faithful view of errors that affect the underlying speech content.

Detailed breakdown of various benchmarks:

Five Modes, One Model

Different applications need different representations of the same speech. A compliance archive may need a verbatim transcript, while a CRM workflow may require normalized text and an analytics system may need an English translation. These transformations are often handled through separate post-processing steps, which add complexity and can introduce additional errors.

Saaras V4 handles these requirements within the model itself. From the same audio input, it can produce five transcript modes: verbatim, normalized, code-mixed, transliterated, and translated.

Verbatim: The exact transcription in the native script of the spoken language
Transcribe: The exact transcription in the native script of the spoken language with numbers and dates normalised
Codemix: Transcribe mode plus keeps native english words in english to improve legibility
Translit: Does transcription but in english script. Colloquially referred to as “WhatsApp language”
Translate: The English translation of the audio with numbers normalized

Because all five come from one model, there are no extra pre-processing steps that could lead to cascading of errors.

Beyond the transcription

a. Best in class Language Identification

Saaras V4 can identify the spoken language directly from the audio and transcribe it in the corresponding native script, without requiring the language to be specified in advance. On verified IndicVoices utterances, language identification error is 5.22% across all 22 Indian languages, falling to 2.9% across the ten most widely spoken languages.

b. Keyterm prompting

Keyterm prompting lets you bias transcription toward words or phrases that are easy to miss, such as product names, people’s names, acronyms, or domain-specific terminology.

Validated as best on ai4bharat/IndicContextEval (Interspeech 2026): Saaras V4 reaches the lowest WER of 16.03% in the L5 (keyword prompting) setting, where the domain-entity list is provided in native script along with the language. For dataset details and competitor scores, see the paper.

c. Built For Real-World Audio

Field audio is compressed, clipped, and full of interference. These are conditions where systems trained on clean benchmarks often fall apart. Saaras V4 is designed to remain reliable under these conditions, including noisy audio, code-mixing, and dialectal variation.

Language	Condition	Format	Transcript
Hindi	Constant wind noise	Verbatim	ए एक बात सुनो ना मैडम दो दो दो दो दो दोनों मिलाकर
Hindi	Distortion and clipping	Transcribe	सर आपको अपनी पॉलिसी के बारे में कुछ पूछना है? मैं आपसे ये पूछ रही थी कि।
Hindi	Constant background	Codemix	होते हो। Nice to meet you brothers. Okay. Thank you. Bye. कुछ कुछ price तो ऐसा है कि जो
Bengali	Heavy traffic	Translit	Maane ami ekta TV kinechilam ami ekta TV installmente niyechilam apnader bank hoyto erokom taka kate okhan theke ki koreche
Kannada	Static noise	Translate	First, in our art, in our culture, first we built a temple. We used to live in huts. We built big, big temples. If you see it, it's a surprise even now. How they must have built it, for a thousand years, for two thousand years, this kind of thing, it's a surprise.
English	Channel distortion	Transcribe	Oh, what a record to read! What a picture to gaze upon! How awful the fact!

d. Low-latency streaming

For real-time voice applications, latency is as important as transcription quality. Saaras V4 supports streaming with time to first token below 150 ms, allowing downstream systems to begin responding with minimal delay and enabling more natural turn-taking.

It also handles long-form audio natively, with multi-minute recordings processed within a second.

Integrating Saaras V4 into your platform

Integrate Saaras v4 into your application with a simple API key from the Sarvam AI dashboard

Choose Your Integration Path:

REST API: Synchronous transcription for audio under 30 seconds
Batch API: Asynchronous processing for files up to 2 hours. Speaker Diarization can be enabled in batch mode
WebSocket (Real-time): Streaming transcription with partial results for live applications

SDKs & Frameworks:

Ready-to-use SDKs for Python 3.9+ and Node.js 18+.

Framework integrations available for Vercel AI SDK, LiveKit Agents and Pipecat Agents

Output Modes:

Configure transcription output: transcribe (default), translate (to English), translit (romanized), or verbatim or codemix

Quick Start:

See the Developer Quickstart for code examples and authentication details

One model across languages

With Saaras V4, you no longer have to trade off English performance against Indic language coverage. It delivers strong performance across both, while supporting low-latency streaming and reliable production workloads. This makes it possible to build multilingual voice applications with one ASR model.