BharatGen

India's AI, built from the ground up

BharatGen is an AI company. We build the data, the research and open-weight AI models for text, speech and vision, so anyone can build on them.

22Indian languages and English in Param2
9consortium institutions, anchored at IIT Bombay
30+research papers at venues such as NeurIPS, ICLR, ACL and CVPR
60+full-time engineers, linguists and builders

Built from scratch. Open weights. Made for India.

Real outputs from our models

Nothing here is staged or edited. Pick a model to see what it can do.

Prompt

Write a binary search program in Python. All comments in the code must be written in Telugu language.

Code · Telugu
Param2 answers
See how it reasoned

User binary search కోసం Python program కోరారు. Comments అన్ని Telugu లో ఉండాలి. Binary search సాధారణ algorithm, array, low, high, mid. Telugu లో వివరణలు ఇవ్వాలి. Code లో comments Telugu లో ఉండాలి, code logic correct ఉండాలి. Python లో function name in English, comments Telugu.

def binary_search(arr, target):
    # ఈ ఫంక్షన్ లక్ష్య విలువను కనుగుకోవడానికి binary search ను అమలు చేస్తుంది
    low = 0
    high = len(arr) - 1

    while low <= high:
        mid = (low + high) // 2

        # మధ్య విలువను పరిశీలిస్తాము
        if arr[mid] == target:
            # లక్ష్య విలువను కనుగొన్నాం, దాని index ను తిరిగి ఇస్తాం
            return mid
        elif arr[mid] < target:
            # లక్ష్య విలువ పెద్దది, కాబట్టి low ను పెంచుతాం
            low = mid + 1
        else:
            # target చిన్నది, కాబట్టి high ను తగ్గిస్తాం
            high = mid - 1

    # లక్ష్య విలువ లేదు, -1 ను తిరిగి ఇస్తాం
    return -1

# ఉదాహరణ వినియోగం
arr = [2, 3, 5, 7, 11, 13]
target = 7
result = binary_search(arr, target)
print(f"లక్ష్య విలువ {target} యొక్క index: {result}")

Real output, not edited. Source: Param2 model card ↗ · More from Param2 →

How we build

Made in India, open to everyone

Our own data, our own tokenisers, our own training. Then we publish the weights.

From scratch

Not fine‑tuned from someone else's model

Pretrained by our team on a corpus we curate with provenance, so we know what each model learned.

Open weights

Download, run and adapt

On Hugging Face and AIKosh, to run on your own machines and fine-tune for your work.

For India

Languages, culture and context

Code-mixed sentences, Indian names, government forms and regional scripts.

Measured honestly

Benchmarks built for India

BhashaBench and ParamBench, open to anyone testing models on Indian knowledge.

Build with BharatGen

Developers, enterprises, government teams and researchers can start with our open weights today.