IndiaAI Mission Releases Three Indigenous AI Models
AI · First reported 3 Nov, 19:02 · Updated 20:16
The IndiaAI Mission released EKA-1 from Soket AI Labs and Dhīr 2B and Dhīr 5B from ZenteiQ. The models differ in scale, language coverage and deployment options, with reported uses ranging from sector-specific applications to offline services.
- EKA-1 has 26.3 billion parameters, was trained on about 7.5 trillion tokens, and supports 22 scheduled Indian languages plus English.
- Dhīr 2B and Dhīr 5B support English and 10 Indian languages, and are designed for on-premise or offline use.
- The models were unveiled in New Delhi on November 3; the IndiaAI Mission supported their development with subsidised compute infrastructure and research-expense support.
Summary written by Lux AI from 2 articles
Key facts
- Models
- EKA-1, Dhīr 2B and Dhīr 5B
- Developers
- Soket AI Labs developed EKA-1; ZenteiQ AiTech Innovations developed Dhīr 2B and Dhīr 5B.
- Launch
- Unveiled in New Delhi on November 3.
- EKA-1 scale
- 26.3 billion parameters; trained on approximately 7.5 trillion tokens.
Source spectrum
Far left
Left
Lean left
Centre
Lean right
Right
Far right
Left 0 · Centre 2 · Right 0 of 2 rated sources
Images from the coverage
Background
- An earlier Lux headline covered a proposed method for measuring AI models and claims about Chinese competition with OpenAI and Anthropic. [3591802e-dcb8-4c4a-b7d0-b5b20af482c6]
- Another earlier Lux headline described research on AI models’ performance when rewriting memes. [a2614655-07f9-4e00-8598-b0f1a74a5d46]
All coverage
Coverage over time
LeftCentreRight
Articles per hour since first reported, by side.
Ownership
- Corporation 100% (2)
Share of the 2 sources by owner type.