AI4Bharat
IIT Madras lab releasing open translation, speech and language models for India
Overview
AI4Bharat is a research lab at IIT Madras that releases open datasets and models for India's 22 scheduled languages, including IndicTrans2 translation, IndicConformer speech recognition, Indic Parler-TTS and the IndicBERT models. Most are MIT or Apache 2.0, though some carry CC-BY-4.0, Llama 2 or GPL-3.0 terms. There is no paid product; models are self-hosted, and the lab collects data for the government's Bhashini mission.
Pricing
Pricing shown for reference only. These figures reflect RECATOOLS research as of 14 Oct 2026 and may be out of date or incomplete. This is not financial or purchasing advice — always confirm the current price on the provider’s official website before making any decision.
Use cases
What you can produce with AI4Bharat
- English-Indic, Indic-English and Indic-Indic translation across 22 scheduled languages with IndicTrans2
- Speech-to-text in Indian languages with IndicConformer and IndicWhisper models
- Natural-sounding Indian-language speech from Indic Parler-TTS and IndicF5
- Romanised text converted to Indian scripts, and back, with IndicXlit
- IndicBERT v3 encoder models for classification and tagging
- Open training corpora including BPCC, IndicVoices, Sangraha and IndicCorpV2
- Translation, speech recognition and speech synthesis models for Bhili
ASEAN Perspective
AI4Bharat in Southeast Asia
Tamil is a co-official language of Singapore and is spoken by roughly 7% of Malaysia's population, making AI4Bharat's Tamil translation and ASR models directly relevant to ASEAN public-sector and edtech deployments. Organisations serving Tamil-speaking communities in Singapore, Malaysia, or Sri Lanka can self-host IndicTrans2 or IndicConformer within their own infrastructure — an advantage for teams with strict data-residency requirements under Singapore's PDPA or Malaysia's PDPA, since no data leaves the host environment. That said, AI4Bharat offers no regional cloud endpoints, SLA guarantees, or ASEAN-specific compliance certifications, so teams requiring enterprise support must pair the models with their own hosting and legal review. The suite does not cover Southeast Asian languages such as Malay, Indonesian, or Filipino.
For anyone building translation, speech recognition or text-to-speech in Indian languages, AI4Bharat's releases are the obvious starting point: IndicTrans2 covers all 22 scheduled languages, including Indian-to-Indian pairs, and the speech models are trained on data the lab collected across more than 400 districts. Most of it is MIT or Apache 2.0, so commercial use is straightforward, but check each model card, because IndicTrans3-beta, Airavata and the Bhili adapters carry other terms.
It is research output, not a service. There is no managed API, uptime commitment or support, and running the models well needs ML engineering. Teams wanting hosted Indic language services should look at Bhashini or Sarvam, both closely tied to this lab.
What we found
AI4Bharat is a research lab at IIT Madras that builds open models and datasets for India's 22 scheduled languages, and it is best read as a supplier of research rather than a product. Its best-known release, IndicTrans2, translates between English and all 22 languages and between Indian languages directly, trained on the Bharat Parallel Corpus Collection of about 230 million sentence pairs, 2.2 million of them made by the lab's own translators. Around it sit IndicConformer and IndicWhisper for speech recognition, Indic Parler-TTS and IndicF5 for speech synthesis, IndicXlit for transliteration and the IndicBERT family, which reached version 3 in December 2025. Document layout parsing and OCR are listed as coming soon.
Licences vary more than this page used to say. IndicTrans2, IndicConformer's per-language models and IndicBERT are MIT, Indic Parler-TTS is Apache 2.0, IndicTrans3-beta is CC-BY-4.0, Airavata carries the Llama 2 licence and some 2026 Bhili translation adapters are GPL-3.0. Most weights sit behind an accept-terms click on Hugging Face.
Funding is public. IIT Madras opened the Nilekani Centre at AI4Bharat in July 2022 with a Rs 36 crore grant from Nilekani Philanthropies, and MeitY funds the lab's data collection as part of the Bhashini mission, across more than 400 districts. Business Today reported that most of the data behind Bhashini comes from AI4Bharat. Lead researcher Mitesh Khapra was named to TIME's 2025 list of the 100 most influential people in AI.
The link to Sarvam AI is close. Sarvam co-founder Pratyush Kumar co-led AI4Bharat, and Sarvam named the lab a collaborator on its IndiaAI Mission models in May 2025; this year the lab fine-tuned Sarvam's translation model for Bhili, its newest language work. There is no paid tier, managed API or support contract: teams self-host, try the demo at models.ai4bharat.org, or use the hosted services on Bhashini, the government platform.
Compiled by RECATOOLS from public sources: vendor documentation and pricing, code repositories and published reporting.
Notable facts
- AI4Bharat researchers visited nearly 500 of India's 700 districts to record speech data covering all 22 official languages across diverse socioeconomic backgrounds.
- The Indian Supreme Court uses AI4Bharat models to translate official legal documents into regional languages.
- Lead researcher Mitesh Khapra appeared alongside Elon Musk and Sam Altman on TIME magazine's 100 Most Influential People in AI 2025 list.
- IndicTrans2 was the first translation model to support all 22 constitutionally scheduled Indian languages, including rare scripts like Ol Chiki (Santali) and Meitei.
Frequently asked questions
About this listing
This entry was compiled from publicly available data including AI4Bharat's official website, press releases, documentation, and reputable third-party publications. RECATOOLS is not affiliated with AI4Bharat unless explicitly stated.
Third-party AI tools update their pricing, features, availability, and policies frequently. Information here may be outdated by the time you read this — we make reasonable efforts to keep listings current, but cannot guarantee absolute accuracy.
For the latest details, please refer to AI4Bharat directly →
Spotted something out of date? Suggest an update →
Related reading
More in LLMs & Chat
