Compare / Nemotron 3.5 Lightning vs Sarvam 30B

Nemotron 3.5 Lightning vs Sarvam 30B

Updated November 2026

Verdict

Nemotron 3.5 Lightning scores higher on the Small LM Index: 63.6 vs 61.6, although their likely ranges overlap, so the lead is not certain. On the 2 benchmarks both labs report, Nemotron 3.5 Lightning wins 2 and Sarvam 30B wins 0. Both need about 22.4 GB of memory at 4-bit quantization. Only Nemotron 3.5 Lightning has a hosted API on OpenRouter ($0.049 per million tokens); the other has to be self-hosted.

Pick Nemotron 3.5 Lightning if…

you want the strongest model overall
you want a hosted API rather than running it yourself
you work with long documents

Pick Sarvam 30B if…

you need a permissive license (Apache-2.0)

Specifications

	Nemotron 3.5 Lightning	Sarvam 30B
Small LM Index	63.6 (likely 55–71)	61.6 (likely 58–65)
Lab	NVIDIA · United States	Sarvam AI · Rest of Asia
Released	2026-08-11	2026-03-03
Parameters	31.6B (3.6B active, MoE)	32.2B (2.4B active, MoE)
Memory at 4-bit	~22.4 GB	~20.8 GB
Context window	1M tokens	131K tokens
Input	text	text
Reasoning	Always thinks	Always thinks
License	OpenMDW-1.1	Apache-2.0
API price / 1M tokens	$0.049 ($0.025 in, $0.12 out)	Self-host only
Hugging Face downloads (30 days)	—	5K

Benchmarks

Higher is better. Bold marks the winner where both labs report a score. ≈ marks the index model's estimate for a score that wasn't reported, ✓ an independent measurement and * a score measured by another lab.

Benchmark	Nemotron 3.5 Lightning	Sarvam 30B
MMLU-Pro Knowledge	81.9	80.0
GPQA Diamond Knowledge	75.4	66.5
AIME Math	≈ 88	88.3
SWE-bench Verified Coding	51.6	≈ 48
τ²-Bench Agentic	—	45.7
MMLU Knowledge	—	85.1
MATH-500 Math	—	97.0
HMMT Math	—	73.3

Scores are mostly reported by each lab with its own settings, so small gaps are noise. See the methodology. Open in the interactive comparison to add more models.

Questions

Which is better, Nemotron 3.5 Lightning or Sarvam 30B?

Which is cheaper to run?

Only Nemotron 3.5 Lightning has a hosted API on OpenRouter ($0.049 per million tokens); the other has to be self-hosted. Both need about 22.4 GB of memory at 4-bit quantization.

Which has the longer context window?

Nemotron 3.5 Lightning has the longer context window: 1M tokens against 131K.

Can I use Nemotron 3.5 Lightning and Sarvam 30B commercially?

Nemotron 3.5 Lightning is released under OpenMDW-1.1 and Sarvam 30B under Apache-2.0. Apache 2.0 and MIT allow commercial use; other licenses can add conditions, so read the license before shipping a product.