# Sarvam LLM

8 Hacker News commenters recommended Sarvam LLM and 8 criticized it, November 2025 to November 2026. It was recommended in 4 of 13 months.

**Verdict:** Indian open-weight language model praised for being trained from scratch, cheap, and strong on local languages. Critics say it hallucinates and lacks transparency about its claims.

## Why people recommend it

- Solid results for a first release (2 comments): "I’ve played around with the 26b parameter model and it’s pretty solid." ([source](https://news.ycombinator.com/item?id=48542590)) "These look like good results for a first model release." ([source](https://news.ycombinator.com/item?id=47286839))
- Accurate answers with few errors (2 comments): "which sarvam got done with 99% accuracy leaving one text error" ([source](https://news.ycombinator.com/item?id=48648076)) "it gave good, unbiased answers, giving a good holistic picture" ([source](https://news.ycombinator.com/item?id=47288748))

## What people warn about

- Hallucinates confidently, lies (2 comments): "it confidently lies / hallucinates without a hint of remorse" ([source](https://news.ycombinator.com/item?id=47288036)) "the hallucinations produced by the model when I poked at it reminded me of model responses before the last couple of years of work" ([source](https://news.ycombinator.com/item?id=47291893))
- Lacks transparency about its claims (2 comments): "It's "open weights" not "open source" and many other (problematic) things" ([source](https://news.ycombinator.com/item?id=47287146)) "The claims are bold. The receipts are missing." ([source](https://news.ycombinator.com/item?id=47075894))
- Weak on domain-specific tasks (2 comments): "Does not handle critical inputs even for moderation tasks" ([source](https://news.ycombinator.com/item?id=47286235)) "Domain specific models simply don’t work. Ever." ([source](https://news.ycombinator.com/item?id=47748829))

## Alternatives hackers recommend

- [Qwen 3.8 27B](https://hackerapproved.com/p/qwen-3-8-27b.md): 1173 recommend
- [Z.ai GLM](https://hackerapproved.com/p/glm-5-2.md): 1043 recommend
- [Kimi](https://hackerapproved.com/p/kimi-k3.md): 931 recommend
- [Gemma 4](https://hackerapproved.com/p/gemma-4.md): 713 recommend
- [Qwen](https://hackerapproved.com/p/qwen.md): 474 recommend

## Cite as

Hacker Approved, "Sarvam LLM", https://hackerapproved.com/p/sarvam-105b/ (counts from 1,795,640 Hacker News comments, November 2025 to November 2026).