Requesty
NVIDIA

nemotron-3-nano-30b-a3b

NVIDIA Nemotron 3 Nano 30B-A3B is a small language MoE model offering high compute efficiency and accuracy for building specialized agentic AI systems. Fully open weights, datasets, and recipes. Text in/out, up to 256K context.

ReasoningTool calling
Try nemotron-3-nano-30b-a3bCompareDocs

Specifications

Context window262K tokens
Max outputβ€”
API typechat
AddedJun 9, 2026
Model IDnvidia/nemotron-3-nano-30b-a3b
Data retentionYes
Used for trainingYes
Provider locationπŸ‡ΊπŸ‡Έ US
Privacy policyNVIDIA Privacy Policy→

Benchmarks

Released 2025-12-15
Coding Indexcoding
19.0%

Artificial Analysis Coding Index β€” a composite of coding evaluations including LiveCodeBench, SciCode and Terminal-Bench.

GPQA Diamondreasoning
75.7%

Graduate-level physics, chemistry & biology questions designed to resist Googling.

Intelligence Indexreasoning
24.3%

Artificial Analysis Intelligence Index β€” a composite of multiple evaluations measuring overall model capability.

Scores are sourced from official model cards, Artificial Analysis, and public leaderboards. Benchmarks measure specific skills and do not capture every aspect of model quality β€” always test on your own workload.

Pricing

Input / 1M
Free
Output / 1M
Free
Cache write
β€”
Cache read
β€”
Estimated cost
100K input + 10K outputFree
1M input + 100K outputFree
10M input + 1M outputFree

Requesty charges exactly what the upstream provider charges β€” no markup, no per-request fees. Prompt caching and smart routing can reduce effective cost by 30-80%.

Quickstart

Drop-in compatible with the OpenAI SDK. Change the base URL, swap in your Requesty API key, and set the model to nvidia/nemotron-3-nano-30b-a3b.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
from openai import OpenAI

client = OpenAI(
    api_key="YOUR_REQUESTY_API_KEY",
    base_url="https://router.requesty.ai/v1",
)

response = client.chat.completions.create(
    model="nvidia/nemotron-3-nano-30b-a3b",
    messages=[
        {"role": "user", "content": "Explain quantum computing in one paragraph."},
    ],
)

print(response.choices[0].message.content)

Other NVIDIA models

nemotron-3-super-120b-a12b
1.0M ctxFree/1M
Coding 31
nemotron-3-ultra-550b-a55b
1.0M ctxFree/1M
Coding 38

Frequently asked questions

How much does nemotron-3-nano-30b-a3b cost?
nemotron-3-nano-30b-a3b is priced at Free per million input tokens and Free per million output tokens when accessed via Requesty. Requesty charges exactly what the upstream provider charges β€” we don't add markup.
What is the context window of nemotron-3-nano-30b-a3b?
nemotron-3-nano-30b-a3b has a context window of 262K tokens. That's roughly 350 words of input you can fit in a single prompt.
How does nemotron-3-nano-30b-a3b perform on benchmarks?
nemotron-3-nano-30b-a3b scores 91.0% on Math Index, 91.0% on AIME 2025, 79.4% on MMLU Pro. See the full benchmark chart above for results across MMLU Pro, GPQA Diamond, SWE-Bench Verified, HumanEval, MATH, AIME, MMMU, and LiveBench.
What can nemotron-3-nano-30b-a3b do?
nemotron-3-nano-30b-a3b supports tool calling, extended reasoning. You can call it through any OpenAI-compatible client by pointing base_url to Requesty.
How do I use nemotron-3-nano-30b-a3b with the OpenAI SDK?
Install the OpenAI SDK, set base_url to "https://router.requesty.ai/v1", set your API key to your Requesty key, and set the model to "nvidia/nemotron-3-nano-30b-a3b". The Quickstart above shows Python, JavaScript and cURL snippets.

Access nemotron-3-nano-30b-a3b through Requesty

One API key, 400+ models, OpenAI-compatible. No markup on provider prices, automatic failover, and smart caching built-in.

Get started freeAll NVIDIA models