nvidia
NVIDIA: Nemotron 3 Nano 30B A3B
nvidia/nemotron-3-nano-30b-a3b
Our take
AI-assisted analysis based on the sources below. Suitability is an editorial assessment; test it on your own workload.
NVIDIA Nemotron 3 Nano 30B A3B is a text-only language model built on a hybrid Mamba-2 and Mixture-of-Experts design with 30B total and about 3.5B active parameters, per NVIDIA's documentation. The vendor describes a configurable reasoning mode that generates reasoning traces before answering, and lists agentic systems, chatbots, RAG, and instruction following as intended uses. Six languages are supported. The model card claims up to 1M tokens of context, while the OpenRouter catalog lists 262,144 tokens plus tool calling and structured outputs.[1][2]
Strengths
- Configurable reasoning: NVIDIA says enabling reasoning traces generally improves answer quality on harder prompts, while thinking can be switched off for direct responses.[2]
- Efficiency-oriented design: a hybrid Mamba-2 and MoE architecture activates only about 3.5B of its 30B parameters per token, which the vendor positions for high compute efficiency.[2]
- Agent-ready feature set: the OpenRouter catalog exposes tool calling, structured outputs, and reasoning parameters, and NVIDIA trained the model for multi-step tool use.[1][2]
Limitations
- Text-only modality: the model accepts and produces text, with no support for image or audio inputs noted in either source.[1][2]
- Evidence is limited to the OpenRouter catalog and NVIDIA's own documentation; all benchmark figures are vendor-reported, so independent validation on target workloads is still needed.[1][2]
- Long-context caveat: the card's 1M token maximum raises VRAM requirements, the default Hugging Face config is 256k, and the OpenRouter listing caps context at 262,144 tokens.[1][2]
Where to use it
- A plausible fit for specialized agentic systems that need function calling, structured outputs, and multi-step tool use, as suggested by both sources.[1][2]
- Could serve general chat, RAG, and instruction-following applications across its six supported languages, subject to workload testing.[2]
- Worth evaluating for coding, math, and science tasks where enabling reasoning traces may improve results, per the vendor's description.[2]
Sources
- [1] OpenRouter catalog · NVIDIA: Nemotron 3 Nano 30B A3BSources checked: Nov 13, 2026
- [2] Model documentation · NVIDIA: Nemotron 3 Nano 30B A3BSources checked: Nov 13, 2026
About this model
Description below is the original text supplied by OpenRouter, shown in the source language.
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...
Specifications
- Model ID
- nvidia/nemotron-3-nano-30b-a3b
- Creator
- nvidia
- Added to OpenRouter
- 14.12.2025
- Context length
- 262,144
- Modality
- text->text
- Input modalities
- text
- Output modalities
- text
- Token pricing
- USD per 1M tokens: in $0.060 / out $0.240
- Knowledge cutoff
- —
- Expiration date
- —
Tasks and capabilities
A grade appears only where the source supplies a matching measurement for that task, the model is active and the cohort holds at least 10 comparable models. The capability column lists facts the source declares — supported formats, reasoning and tool parameters — which are not measured quality.
| Task | Grade | Source-declared capability |
|---|---|---|
| General quality | No measurement | reasoning: optional · parameters: include_reasoning, reasoning |
| Coding | CC — below the B band | — |
| Agents | CC — below the B band | parameters: structured_outputs, tool_choice, tools |
| Writing | No measurement | output: text |
| Reasoning | No measurement | reasoning: optional · parameters: include_reasoning, reasoning |
| Document analysis | No measurement | — |
| Vision | No measurement | — |
| Image generation | No measurement | — |
| Audio | No measurement | — |
| Video | No measurement | — |
| Embeddings | No measurement | — |
Evaluations
| Source | Benchmark | Value | Grade | Data retrieved |
|---|---|---|---|---|
| Artificial Analysis | Coding index | 14.4 | CC — below the B band | 13.11.2026 |
| Artificial Analysis | Agentic index | 1 | CC — below the B band | 13.11.2026 |
How grades work
Each letter grade compares a model only against other models scored by the same source and metric in the current snapshot. S means the top 10%, A the next 20%, B the next 30%, C the rest. Models without a published measurement are shown as not rated — a supported format alone never implies measured quality.
- S — S — top 10% of the cohort
- A — A — next 20% of the cohort
- B — B — next 30% of the cohort
- C — C — below the B band
Providers
Optional live provider statistics from OpenRouter. Missing values are unknown, not zero.
| Provider | Context length | Price in /1M | Price out /1M | Latency p50 (30m, s) | Throughput p50 (30m, tok/s) | Uptime (30m) |
|---|---|---|---|---|---|---|
| Nebiusfp8 | 262,144 | $0.060 | $0.240 | — | — | 100% |
Provider data retrieved: 13.11.2026
Related articles
No published articles identify this model yet.