Nemotron 3 Nano 30B A3B
NVIDIA · released Dec 14, 2025
NVIDIA Nemotron 3 Nano 30B A3B is a small language MoE model with highest compute efficiency and accuracy for developers to build specialized agentic AI systems. The model is fully...
Specification
- Context window
- 262K
- Max output
- 236K
- Knowledge cutoff
- Not stated
- Parameters
- 31.6B
- Licence
- other
- Serving providers
- 2
- Moderated
- No
- Uptime
- 100.0%
Intelligence
6.8
21th percentile
Coding
—
Coding Index
Agentic
1.0
Agentic Index
Output speed
201 t/s
Median across providers
Latency
362ms
Time to first token
Cost per task
$0.02
Estimated
Benchmarks
Where the score comes from
The Intelligence Index is a composite. These are the underlying evaluations this model was actually measured on.
Evaluation scores
Percentage correct · higher is better
- MMLU-Pro57.9%
- GPQA Diamond39.9%
- IFBench37.5%
- LiveCodeBench36.0%
- τ²-bench (Telecom)25.4%
- AIME 202513.3%
- Terminal-Bench Hard12.1%
- AA-LCR (long context)10.7%
- Humanity's Last Exam4.6%
An evaluation missing from this list was not run for this model — it is not a zero.
View as table
| Evaluation | Score |
|---|---|
| MMLU-Pro | 57.9% |
| GPQA Diamond | 39.9% |
| IFBench | 37.5% |
| LiveCodeBench | 36.0% |
| τ²-bench (Telecom) | 25.4% |
| AIME 2025 | 13.3% |
| Terminal-Bench Hard | 12.1% |
| AA-LCR (long context) | 10.7% |
| Humanity's Last Exam | 4.6% |
Against its peers
Intelligence Index · this model highlighted, nearest peers in grey
- Nemotron 3 Nano 30B A3BNVIDIA6.8
- Mistral Large 2407Mistral AI6.8
- GLM 4.5VZ.ai6.7
- Gemini 2.5 Flash LiteGoogle6.7
- 6.7
- GPT-4o-miniOpenAI6.7
- GPT-4o-mini (2024-07-18)OpenAI6.7
- GPT-4OpenAI6.7
Peers are the models sitting closest on the Intelligence Index — the set you would realistically choose between.
View as table
| Model | Intelligence |
|---|---|
| Nemotron 3 Nano 30B A3B | 6.8 |
| Mistral Large 2407 | 6.8 |
| GLM 4.5V | 6.7 |
| Gemini 2.5 Flash Lite | 6.7 |
| Qwen2.5 Coder 32B Instruct | 6.7 |
| GPT-4o-mini | 6.7 |
| GPT-4o-mini (2024-07-18) | 6.7 |
| GPT-4 | 6.7 |
Percentile among all indexed models
Pricing
What it costs to run
List prices per million tokens, plus what one representative task works out to.
List price
- Input / 1M tokens
- $0.06
- Output / 1M tokens
- $0.24
- Cached input / 1M
- Not offered
- Blended 3:1
- $0.105
One task, estimated
$0.02
- Input tokens
- 50,000
- Output tokens
- 80,000
- Profile
- Reasoning
Estimated from list pricing: 50K input tokens plus 80K output tokens for reasoning models (25K for non-reasoning).