Ministral 3 3B vs NVIDIA Nemotron Nano 9B v2

Prices checked

Ministral 3 3B takes 3 of 5 rounds against NVIDIA Nemotron Nano 9B v2. It is also the cheaper of the two to run, at $0.04 per million tokens against $0.07.

Mistral AIWins 3 rounds

Ministral 3 3B

131k context · tool calling · reads images · open weights

$0.04/M tokens

Cheapest price, at Azure AI Foundry · 2.6¢ per 1,000 pages

NVIDIA

NVIDIA Nemotron Nano 9B v2

131k context · tool calling · open weights

$0.07/M tokens

Cheapest price, at DeepInfra · 3.4¢ per 1,000 pages

Maker's price in (lower wins)

$0.10/M

$0.06/M

Maker's price out (lower wins)

$0.10/M

$0.23/M

Cheapest price (lower wins)

$0.04/M at Azure AI Foundry

$0.07/M at DeepInfra

Maker's price, blended (lower wins)

$0.10/M

$0.10/M

Cost per 1,000 pages (lower wins)

2.6¢

3.4¢

Context

131k tokens

131k tokens

Open weights

yes

yes

Tool calling

yes

yes

Structured output

yes

yes

Reads images

yes

no

Reasoning

no

no

Released

2 Dec 2025

18 Aug 2025

Providers selling it

18

5

Price we measure against

the maker's own

Amazon Bedrock

Trains on your prompts (no is better)

not published

not published

Providers selling both · 2

  • Eden AI$0.10 / $0.10
  • Fireworks AI$0.10 / $0.20

Ministral 3 3B, then NVIDIA Nemotron Nano 9B v2, per million tokens.

Go further

Add a third or fourth model, or change the weighting, in the tool.

Frequently asked

Which is cheaper, Ministral 3 3B or NVIDIA Nemotron Nano 9B v2?
Ministral 3 3B, at $0.04 per million tokens against $0.07 for NVIDIA Nemotron Nano 9B v2, blended 3 in : 1 out, read on 21 Oct 2026.
Where is each one cheapest?
Ministral 3 3B: $0.04 at Azure AI Foundry. NVIDIA Nemotron Nano 9B v2: $0.07 at DeepInfra.
Can I buy both from one provider?
Yes. 2 providers sell both: Eden AI, Fireworks AI. The cheapest of them for the two together is Eden AI.
Which holds more context?
Both hold 131k tokens in one call.

Prices exclude tax, read 21 Oct 2026 from each provider's own catalog. They are per M tokens, blended 3 in : 1 out, and a page is 600 tokens in and 60 out. Bars are scaled to the larger value in each row.

More