---
title: "Soket AI models: release history and timeline"
description: "Soket AI is a Bengaluru research lab building the Project EKA foundation models for math, code and reasoning in Indian and Global South languages, along with Indic language and speech models."
type: "release-archive"
canonical_url: "https://theaiengineering.com/releases/soket-ai"
markdown_url: "https://theaiengineering.com/releases/soket-ai.md"
author: "The AI Engineering Editors"
publisher: "The AI Engineering"
updated: "2026-11-03"
citation: "The AI Engineering Editors. \"Soket AI models: release history and timeline.\" The AI Engineering, updated Nov 03, 2026. https://theaiengineering.com/releases/soket-ai"
license: "CC BY 4.0"
license_url: "https://creativecommons.org/licenses/by/4.0/"
data_csv: "https://theaiengineering.com/releases.csv"
data_json: "https://theaiengineering.com/releases.json"
---

# Soket AI models: release history and timeline

Soket AI is a Bengaluru research lab building the Project EKA foundation models for math, code and reasoning in Indian and Global South languages, along with Indic language and speech models.

- Headquarters: Bengaluru, India
- Website: https://soket.ai

4 releases from 1 lab, 2 with open weights. Dates are the labs' official announcement dates.

## November 2026

### EKA-1

EKA-1 is the first model from Soket AI's Project EKA, a 26B-parameter sparse hybrid model trained from scratch for math, coding, reasoning and agentic work across Indian, Global South and programming languages.

- Date: 2026-11-03
- Lab: Soket AI
- Type: LLM
- Weights: Proprietary
- Modalities: text → text
- Context window: 64,000 tokens
- Parameters: 26B
- Available at launch: Enterprise
- Permalink: https://theaiengineering.com/releases/2026/11#eka-1

Highlights:

- Soket's RALSNet design stacks 52 layers that mix attention, recurrent and expert layers
- Trained on 7.5 trillion tokens with an in-house tokenizer
- Supports tool calling and Soket's open-source Loop agent framework
- At launch Soket listed a 30.49 GB checkpoint on IndiaAI's AIKosh under Apache 2.0 with restricted visibility and asked developers to apply for early access; it also offered on-premises deployment through its inference platform

Source: [EKA-1-26B](https://aikosh.indiaai.gov.in/home/models/details/eka_1_26b_1.html)

## January 2026

### Sarthi-Agri-v1

Sarthi-Agri-v1 is a research-preview model from Soket AI, fine-tuned from Google's Gemma 3 27B to generate structured crop advisories for farmers in Hindi and English.

- Date: 2026-01-17
- Lab: Soket AI
- Type: LLM
- Weights: Open weights
- Modalities: text → text
- License: Apache-2.0
- Available at launch: Open weights
- Permalink: https://theaiengineering.com/releases/2026/01#sarthi-agri-v1

Highlights:

- Trained with LoRA supervised fine-tuning on Soket's Sarathi-AgriData dataset
- Uses control tokens to separate internal reasoning from the final advisory
- Takes structured inputs such as crop, month, soil type, region and weather
- Soket noted the model does not replace certified agricultural experts

Source: [soketlabs/sarthi-agri-v1 · Hugging Face](https://huggingface.co/soketlabs/sarthi-agri-v1)

## November 2025

### Dhrith

Dhrith is Soket AI's speech recognition model for Hindi-English code-mixed speech that adds tags for emotion and tone to its transcripts.

- Date: 2025-11-06
- Lab: Soket AI
- Type: Speech
- Weights: Proprietary
- Modalities: audio → text
- Permalink: https://theaiengineering.com/releases/2025/11#soket-dhrith

Highlights:

- Soket reported 11.19% WER on its Hindi-English benchmark, second only to Gemini 2.5 Flash among the systems it tested
- Marks expressive cues such as tone and emotion inline with the transcript
- Soket said expression tagging degraded in noisy audio
- Soket said public APIs and developer access would follow

Source: [Introducing Dhrith: Emotionally Intelligent ASR](https://soket.ai/blogs/dhrith)

## April 2024

### Pragna-1B

Pragna-1B is Soket AI's open 1.25B-parameter decoder-only model for Hindi, Gujarati, Bangla and English, designed for on-device use.

- Date: 2024-04-30
- Lab: Soket AI
- Type: Small model
- Weights: Open weights
- Modalities: text → text
- Context window: 2,048 tokens
- Parameters: 1.25B
- License: Apache-2.0
- Available at launch: Open weights
- Permalink: https://theaiengineering.com/releases/2024/04#pragna-1b

Highlights:

- Architecture follows TinyLlama, with 22 layers, grouped-query attention and a 69,632-token vocabulary
- Custom tokenizer that Soket reported at 2.85 tokens per Kannada word versus 5.81 for Gemma-7B
- Continued pretraining on about 150 billion tokens focused on Hindi, Bangla and Gujarati
- Base weights went to Hugging Face, while an instruction-tuned version ran as a research preview chat

Source: [Introducing Pragna-1B: Soket AI Labs' Multilingual LLM](https://soket.ai/blogs/pragna_1b)

## Cite and reuse this data

These release records are free to reuse under CC BY 4.0, including commercially. Credit The AI Engineering and link to this page's URL. Dates are the labs' official announcement dates, and every record links to its source. List fields in the CSV are separated by " | ".

The AI Engineering Editors. "Soket AI models: release history and timeline." The AI Engineering, updated Nov 03, 2026. https://theaiengineering.com/releases/soket-ai

HTML link:

```html
<a href="https://theaiengineering.com/releases/soket-ai">Soket AI models: release history and timeline</a> (The AI Engineering)
```

- Title: Soket AI models: release history and timeline
- Author: The AI Engineering Editors
- Publisher: The AI Engineering
- Updated: 2026-11-03
- URL: https://theaiengineering.com/releases/soket-ai
- Markdown: https://theaiengineering.com/releases/soket-ai.md
- License: CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/)
- Full dataset as CSV: https://theaiengineering.com/releases.csv
- Full dataset as JSON: https://theaiengineering.com/releases.json
