Ministral 3 (8B Instruct 2512) vs Nemotron 3 Ultra (550B A55B)

Comparison

As of August 25, 2026, Ministral 3 (8B Instruct 2512) is 89% cheaper per million tokens than Nemotron 3 Ultra (550B A55B). Ministral 3 (8B Instruct 2512) is $0.15 / $0.15 per million input/output tokens versus Nemotron 3 Ultra (550B A55B) at $0.50 / $2.20. Source: https://openrouter.ai/api/v1/models

Source: OpenRouter pricing (https://openrouter.ai/api/v1/models)

Better value

Ministral 3 (8B Instruct 2512)

Mistral

$0.30blended / 1M

Input

$0.15

Output

$0.15

262.1K ctx|Open Source

89%

cheaper

Nemotron 3 Ultra (550B A55B)

NVIDIA

$2.70blended / 1M

Input

$0.50

Output

$2.20

1M ctx|Open Source

Save $2.40 per million tokens by choosing Ministral 3 (8B Instruct 2512) over Nemotron 3 Ultra (550B A55B).

Ministral 3 (8B Instruct 2512) vs Nemotron 3 Ultra (550B A55B) comparison

Token pricing, specs, and benchmarks side by side

Metric

Ministral 3 (8B Instruct 2512)

Nemotron 3 Ultra (550B A55B)

Provider

Mistral

NVIDIA

License

Open Source

Release Date

2025-12-04

2026-06-04

Pricing (per 1M tokens)

Input Price-70%

$0.15

$0.50

Output Price-93%

$0.15

$2.20

Blended (1M + 1M)

$0.30

$2.70

Model Details

Context Window

262.1K

Max Output Tokens

N/A

Knowledge Cutoff

N/A

2025-09-30

Throughput

N/A

Benchmarks

HLE

—

37.4%

GPQA

—

87%

GPQA Diamond

—

86.1%

SWE-bench Verified

—

70.7%

LiveCodeBench

—

86.0%

LiveCodeBench v6

—

89%

SciCode

—

44.6%

Vibe Code Bench

—

7.6%

BrowseComp

—

44.4%

IFBench

—

81.7%

AA-LCR

—

65.4%

FinanceAgent v2

—

37.5%

SWE-bench Multilingual

—

67.7%

Terminal-Bench 2.1

—

56.4%

GDPval

—

46.7%

FinanceAgent v1.1

—

53.7%

MATH

87.6%

—

Apex

—

84.8%

Arena-Hard

50.9%

—

Code Migration

—

4.9%

CorpFin v2

—

65.5%

CritPt

—

3.1%

EMB

—

31.7%

Finance Agent

—

53.7%

Imo Answerbench

—

92.3%

Longbench V2

—

61.9%

MedCode

—

38.6%

Mm Mt Bench

8.1%

—

MMLU-Pro

—

86.8%

Mmlu Prox

—

83%

Multichallenge

—

63.8%

Omniscience

—

78.7%

Pinchbench

—

90%

Profbench

—

56%

ProofBench

—

Ruler

—

94.7%

Tau3 Bench

—

22.6%

Tax Eval v2

—

73.1%

Wild Bench

66.8%

—

Wmt24

—

83.7%

Verdict

Ministral 3 (8B Instruct 2512) vs Nemotron 3 Ultra (550B A55B): the bottom line

Both models are closely matched on benchmarks. Ministral 3 (8B Instruct 2512) has the pricing advantage, making it the better value unless Nemotron 3 Ultra (550B A55B)'s specific capabilities are critical to your workflow.

View Ministral 3 (8B Instruct 2512) details View Nemotron 3 Ultra (550B A55B) details

Ministral 3 (8B Instruct 2512) vs Nemotron 3 Ultra (550B A55B) FAQ

Which is cheaper, Ministral 3 (8B Instruct 2512) or Nemotron 3 Ultra (550B A55B)?

As of August 25, 2026, Ministral 3 (8B Instruct 2512) is 89% cheaper on blended LLM API pricing. Ministral 3 (8B Instruct 2512) costs $0.15 per million input tokens and $0.15 per million output tokens ($0.30 blended 1M-in + 1M-out). Nemotron 3 Ultra (550B A55B) costs $0.50 / $2.20 per million tokens ($2.70 blended). Sources: https://openrouter.ai/api/v1/models and https://openrouter.ai/api/v1/models. The cheapest LLM for your app still depends on how many output tokens you generate.

How much does Ministral 3 (8B Instruct 2512) vs Nemotron 3 Ultra (550B A55B) cost per million tokens?

Ministral 3 (8B Instruct 2512) Mistral API pricing is $0.15 input and $0.15 output per million tokens via OpenRouter. Nemotron 3 Ultra (550B A55B) OpenRouter API pricing is $0.50 input and $2.20 output per million tokens via OpenRouter. Use those four numbers, not ChatGPT Plus or Claude Pro subscription prices, when you are comparing APIs.

Which is better for coding, Ministral 3 (8B Instruct 2512) or Nemotron 3 Ultra (550B A55B)?

Nemotron 3 Ultra (550B A55B) has a published SWE-bench Verified score of 70.7%. Ministral 3 (8B Instruct 2512) does not have that eval on this page yet.

What is the context window for Ministral 3 (8B Instruct 2512) vs Nemotron 3 Ultra (550B A55B)?

Ministral 3 (8B Instruct 2512) supports a 262K token context window, while Nemotron 3 Ultra (550B A55B) supports 1M tokens. Nemotron 3 Ultra (550B A55B) offers a larger context window. Filling a larger window bills more input tokens, so the cheaper-per-million model can still cost more on long documents.

Which model performs better on benchmarks, Ministral 3 (8B Instruct 2512) or Nemotron 3 Ultra (550B A55B)?

Shared benchmark scores for Ministral 3 (8B Instruct 2512) and Nemotron 3 Ultra (550B A55B) are still limited. Check the comparison table for the evals each model has published. Check the comparison table for GPQA Diamond, SWE-bench, MMLU, HLE, and the other evals we track. Token price and benchmark score together are the usual LLM comparison, not either number alone.

Is Ministral 3 (8B Instruct 2512) or Nemotron 3 Ultra (550B A55B) better for production use?

Both Ministral 3 (8B Instruct 2512) and Nemotron 3 Ultra (550B A55B) are production API models. For cost-sensitive production traffic, Ministral 3 (8B Instruct 2512) has better economics on the blended 1M-in + 1M-out scale. For maximum capability, weight the benchmark table for your domain. Many production stacks route cheap models for drafts and a frontier model for the hard turn, which is usually cheaper than sending everything to the expensive API.

Can I switch between Ministral 3 (8B Instruct 2512) and Nemotron 3 Ultra (550B A55B) in my app?

Yes. Call Ministral 3 (8B Instruct 2512) through OpenRouter and Nemotron 3 Ultra (550B A55B) through OpenRouter with separate API keys, or through a gateway that already wraps both. Keep prompts in tokens, not characters, when you estimate the invoice. AnotherWrapper templates can swap providers without rewriting auth and billing.

How accurate is this Ministral 3 (8B Instruct 2512) vs Nemotron 3 Ultra (550B A55B) pricing data?

List rates are the latest published Mistral API pricing and OpenRouter API pricing, quoted per million tokens. Official rate sheets: https://openrouter.ai/api/v1/models and https://openrouter.ai/api/v1/models. Prompt caching, batch APIs, and committed-use discounts can change the invoice. Benchmark scores are from official publications and independent evals, updated as new numbers land.

What is the output token limit for Ministral 3 (8B Instruct 2512) and Nemotron 3 Ultra (550B A55B)?

Ministral 3 (8B Instruct 2512) supports up to an unspecified number of output tokens per request, while Nemotron 3 Ultra (550B A55B) supports up to an unspecified number of output tokens. Output tokens are usually the expensive half of API pricing, so a higher max output cap is a capability, not a discount.

Which model has better throughput, Ministral 3 (8B Instruct 2512) or Nemotron 3 Ultra (550B A55B)?

Neither Ministral 3 (8B Instruct 2512) nor Nemotron 3 Ultra (550B A55B) has published throughput data on this page yet. Throughput (tokens per second) and time-to-first-token are why two models with the same per-million-token price can still feel different in chat and agents.

How do I switch the models in this Ministral 3 (8B Instruct 2512) vs Nemotron 3 Ultra (550B A55B) comparison?

Use the model pickers at the top of this LLM comparison. Search any model in the index and the URL updates. Open Ministral 3 (8B Instruct 2512) or Nemotron 3 Ultra (550B A55B) from the sidebar for the single-model API pricing page, or go back to the LLM pricing table to sort by cheapest blended cost.

Ministral 3 (8B Instruct 2512) vs Nemotron 3 Ultra (550B A55B) pricing

This comparison covers API pricing, context window, throughput, and benchmark scores. Ministral 3 (8B Instruct 2512) costs $0.15/M input and $0.15/M output via OpenRouter. Nemotron 3 Ultra (550B A55B) costs $0.50/M input and $2.20/M output via OpenRouter.