Sarvam 30B (Reasoning)

Sarvam·Released 2026-03-06

Compare

Model Family

Sarvam 105B (Reasoning)2026-03-06 Sarvam M (Reasoning)2025-05-23

Quality Index

12.3

322nd of 442

Top 74%

Coding Index

7.9

282nd of 352

Top 80%

Price/1M

$0.00

1st cheapest

100% below median

Top 27%

Speed

305 tok/s

Top 2%

TTFT

1.57s

Market Position

Sarvam 30B (Reasoning)Market Average

Pricing

Input

$0.00

per 1M tokens

Output

$0.00

per 1M tokens

Blended

$0.00

per 1M tokens

Cheaper than 73% of models. Median price is $0.31/1M tokens.

Cost Calculator

Tokens per day1M

100K100M

Daily

$0.00

Monthly

$0.00

vs. Similar Models

Claude 3 HaikuQ:0.0

$0.50

DeepSeek-V2.5Q:0.0

$0.00

Gemini 2.0 Flash Thinking Experimental (Dec '24)Q:0.0

$0.00

Olmo 3.1 32B InstructQ:-0.1

$0.30

Performance

305

tokens/sec

Faster than 98% of models

1.57

seconds

Faster than 16% of models

8.12

seconds

Faster than 21% of models

Market Median

46 tok/s

569% faster

Median TTFT

0.42s

275% slower

Speed Comparison

gpt-oss-20B (high)

304 tok/s-0%

gpt-oss-120B (low)

299 tok/s-2%

Gemini 2.5 Flash-Lite (Reasoning)

317 tok/s+4%

Benchmarks

MMLU-ProNot evaluated

GPQA Diamond

63.3%

HLE

7.0%

LiveCodeBenchNot evaluated

SciCode

19.2%

TerminalBench Hard

2.3%

MATH-500Not evaluated

AIMENot evaluated

AIME 2025Not evaluated

IFBench

26.5%

Long Context Recall

0.0%

Tau2

34.5%

Market AverageTop Score

Quick Compare

Similar Models

Claude 3 Haiku

Anthropic

Q: 12.3$0.50/1M200K ctx

Slower: 56%

Gemini 2.0 Flash Thinking Experimental (Dec '24)

Google

Q: 12.3$0.00/1M

Slower: 100%

DeepSeek-V2.5

DeepSeek

Q: 12.3$0.00/1M

Slower: 100%

Olmo 3.1 32B Instruct

Allen AI

Q: 12.2$0.30/1M66K ctx

Slower: 82%

Llama 3.1 Instruct 70B