Sarvam AI Releases Saaras V4: A Speech-to-Text Model for All 22 Indian Languages and Global English

Sarvam AI released Saaras V4, a speech-to-text model covering all 22 scheduled Indian languages plus global English accents, available today through Sarvam's API as model="saaras:v4". Sarvam reports state-of-the-art accuracy across all 22 languages and, on the Kathbath Noisy set measured with LLM-WER, an error rate under half that of Deepgram Nova-3 and GPT-4o Transcribe, though all figures are vendor-reported with no independent reproduction published. The model pairs an audio encoder with a temporal-downsampling adapter and Sarvam-3B, a 3B-parameter hybrid state-space decoder, and adds five output modes plus keyterm prompting of up to 50 terms of 64 characters each; weights are not public and SageMaker self-hosting docs cover Saaras v3 only.

Sarvam AI https://www.sarvam.ai/ has released Saaras V4 https://www.sarvam.ai/blogs/introducing-saaras-v4 , the newest generation of its speech recognition model. It covers all 22 scheduled Indian languages plus English, now including global English accents. Sarvam reports state-of-the-art accuracy across all 22 languages. Is it deployable? Yes, through Sarvam’s API today, using model="saaras:v4" . Weights are not public, and Sarvam’s SageMaker self-hosting docs https://docs.sarvam.ai/api/self-hosted/sagemaker/deploy-saaras currently cover Saaras v3 only. What is Inside Saaras V4 Saaras V4 is an encoder-decoder system. An audio encoder converts the waveform into embeddings that carry phonetic and acoustic detail. A temporal-downsampling adapter then shortens that sequence and projects it into the language model’s embedding space. This keeps long recordings inside the decoder’s context budget. The decoder is Sarvam-3B, a 3B-parameter hybrid state-space language model trained from scratch in-house. It reads the audio features alongside a text prompt. It then emits the transcript autoregressively, feeding each token back as input for the next. Benchmark Results - English : Sarvam evaluated 7 English datasets. Six come from Hugging Face’s Open ASR Leaderboard https://huggingface.co/datasets/hf-audio/open-asr-leaderboard : AMI, GigaSpeech, LibriSpeech clean, LibriSpeech other, SPGISpeech and VoxPopuli. The seventh is AI4Bharat’s Indian-accented Svarah https://huggingface.co/datasets/ai4bharat/Svarah . Scoring follows the leaderboard’s normalization code https://github.com/huggingface/open asr leaderboard . Saaras V4 posts the lowest average WER among the models Sarvam benchmarked. - Indic : On Vistaar https://github.com/AI4Bharat/vistaar , Sarvam reports results across 10 Indian languages using both WER and LLM-WER https://www.sarvam.ai/blogs/evaluating-indian-language-asr . LLM-WER adds a semantic check. It separates real meaning errors from harmless spelling or formatting variants common in Indic scripts. - Noisy audio : On Kathbath Noisy, measured with LLM-WER, Sarvam says Saaras V4’s error rate is under half that of Deepgram Nova-3 and GPT-4o Transcribe. The set includes compressed, clipped and background-heavy recordings. - Language ID : On verified IndicVoices utterances, language identification error is 2.9% across the top 10 Indian languages. It is 5.22% across all 22. It is important to note that all numbers above are vendor-reported. Independent reproduction has not been published yet. 5 Output Modes From 1 Model The same audio can return 5 representations, selected through the mode parameter: - transcribe default : native script with numbers and dates normalized. - verbatim : every word as spoken, fillers and spoken numbers kept. - codemix : native script, with English words left in English. - translit : the full utterance in Latin script. - translate : an English translation with numbers normalized. Sarvam’s argument is simple. Handling these inside the model removes post-processing steps that can compound errors. Keyterm Prompting Keyterm prompting https://docs.sarvam.ai/api/api-guides-tutorials/speech-to-text/how-to/keyterms is new in V4 and works only with saaras:v4 . You pass a JSON list under keyterms , with up to 50 terms of 64 characters each. Keyterms bias recognition; they do not guarantee output. Use codemix mode when a brand such as PhonePe must stay in Latin script. On IndicContextEval https://huggingface.co/datasets/ai4bharat/IndicContextEval paper https://arxiv.org/pdf/2606.19157 , Interspeech 2026 , Saaras V4 reports 16.03% WER in the L5 keyword-prompting setting. Sarvam says that is the lowest score on the benchmark. Streaming, Long Audio and Pricing - Streaming: WebSocket https://docs.sarvam.ai/api/api-guides-tutorials/speech-to-text/realtime-streaming with partial results and time to first token below 150 ms. - REST: synchronous https://docs.sarvam.ai/api/api-guides-tutorials/speech-to-text/rest-api transcription for clips up to 30 seconds. - Batch: asynchronous https://docs.sarvam.ai/api/api-guides-tutorials/speech-to-text/batch-api jobs up to 2 hours per file, with optional speaker diarization. - SDKs: Python 3.9+ and Node.js 18+, plus LiveKit Agents https://docs.livekit.io/agents/models/stt/sarvam/ , Pipecat https://docs.pipecat.ai/api-reference/server/services/stt/sarvam and Vercel AI SDK https://docs.sarvam.ai/api/integration/vercel-ai-sdk integrations. - Price: Sarvam lists speech-to-text at ₹30 per hour https://www.sarvam.ai/api-pricing for real-time, streaming and batch, and ₹45 per hour with diarization. Saaras v3 stays the default model https://docs.sarvam.ai/api/getting-started/models/saaras . V4 uses the same request shape, so switching is a 1-line change. Saaras V4 vs Closest Competitors These are the 3 systems Sarvam benchmarked against. Figures come from each vendor’s public docs and pricing pages, checked on September 26, 2026. | Feature | Sarvam Saaras V4 | Deepgram Nova-3 | ElevenLabs Scribe v2 | OpenAI GPT-4o Transcribe | |---|---|---|---|---| | Indian scheduled languages of 22 | 22 https://docs.sarvam.ai/api/getting-started/models/saaras | 11 https://developers.deepgram.com/docs/models-languages-overview | 14 https://elevenlabs.io/docs/overview/capabilities/speech-to-text | Not listed per language https://developers.openai.com/api/docs/guides/speech-to-text | | Total languages | 23 22 Indian + English | 45+ | 90+ | Multilingual | | Keyterm biasing | Up to 50 terms | Yes, paid add-on https://deepgram.com/pricing | Up to 1,000 batch , 50 realtime , paid add-on | Free-text prompt | | Built-in output modes | 5 transcribe, verbatim, codemix, translit, translate | Transcript plus Smart Formatting | Verbatim or no verbatim | Transcript | | Real-time streaming | WebSocket, under 150 ms TTFT vendor claim | Yes WebSocket | Scribe v2 Realtime, about 150 ms | File streaming; live via Realtime API | | Speaker diarization | Batch API | Yes | Up to 32 speakers | Separate gpt-4o-transcribe-diarize model | | List price | ₹30/hour https://www.sarvam.ai/api-pricing | $0.0052/min https://deepgram.com/pricing multilingual, pre-recorded | $0.22/hour https://elevenlabs.io/pricing/api batch | ~$0.006/min https://developers.openai.com/api/docs/pricing | | Self-hosting | Not for V4 yet v3 on SageMaker | Yes https://developers.deepgram.com/docs/self-hosted-introduction | Cloud API | Cloud API | Key Takeaways - Saaras V4 covers all 22 Indian languages plus global English in 1 model. - A 3B hybrid state-space decoder, trained from scratch, sits behind an audio encoder. - Keyterm prompting accepts up to 50 terms and scored 16.03% WER on IndicContextEval L5. - 5 output modes and sub-150 ms streaming TTFT come from the same model. - API-only today at ₹30 per hour; self-hosting docs still cover v3. Check out the Technical Details https://www.sarvam.ai/blogs/introducing-saaras-v4 . All credit goes to the researcher of this project. Also, feel free to follow us on Twitter https://x.com/intent/follow?screen name=marktechpost and don’t forget to join our 150k+ML SubReddit https://www.reddit.com/r/machinelearningnews/ and Subscribe to our Newsletter https://magic.beehiiv.com/v1/f5e63dd4-5653-4f09-83e2-321a8b1ba526?email={{email}} . Wait are you on telegram? now you can join us on telegram as well. https://t.me/machinelearningresearchnews Need to partner with us for promoting your GitHub Repo OR Hugging Face Page OR Product Release OR Webinar etc.? Connect with us https://forms.gle/MJjjVDPS7whH8Ngs6 Asif Razzaq is the CEO of Marktechpost AI Media Inc.. As a visionary entrepreneur and engineer, Asif is committed to harnessing the potential of Artificial Intelligence for social good. His most recent endeavor is the launch of an Artificial Intelligence Media Platform, Marktechpost, which stands out for its in-depth coverage of machine learning and deep learning news that is both technically sound and easily understandable by a wide audience. The platform boasts of over 2 million monthly views, illustrating its popularity among audiences.