Nemotron 3.5 Lightning 30B A3B

The small text model of NVIDIA's Nemotron 3.5 line, a 30B hybrid of Mamba-2, MoE and attention layers activating 3B per token, under the OpenMDW 1.1 license.

Params: 30B, 3B active
Context: 1M tokens
Cutoff: Sep 2025
Released: Aug 2026
License: OpenMDW 1.1

Visit Nvidia Compare vs other LLMs →

Hosted API price

Cheapest via Azure · per 1M tokens

USD

Input: $0.06 /1M
Output: $0.22 /1M
Hosted by: 2 providers

Hosted API pricing

How we price

Order

Providers are listed cheapest input rate first, then by output rate, a tie going to the creator's own listing. A provider that lists the model without a published rate sits at the end.

Prices

Base-tier, on-demand rates in USD per 1M tokens, converted at ECB reference rates where a provider publishes in another currency. Cached-input and batch rates show only where a provider publishes them. "Cheapest" marks the one provider with the lowest input and output pair; a tie wears no badge.

Cost

Input rate times the input tokens plus output rate times the output tokens, at the monthly volume set above the table. Cached-input, batch, long-context and reasoning-token billing are not modelled.

Transparency and funding

Affiliates: Affiliate links are marked. We may earn a commission if you click them, but commissions never affect the order.

Every provider serving Nemotron 3.5 Lightning 30B A3B, cheapest input first. Set your monthly volume to see what each would bill.

Input M Output M

Every provider serving Nemotron 3.5 Lightning 30B A3B, with its price per 1M input and output tokens and what the volume above would cost
Provider	Input /1M	Output /1M	Cached input /1M	Cost at 10M in + 2M out	Link
Azure	$0.06	$0.22	$0.01	$1.04	Visit website
Nvidia Creator	Open weights, no hosted price				Visit website

Heads up: Base-tier, on-demand rates per 1M tokens; cached, batch and long-context tiers excluded. A provider may serve a shorter context or a quantized build than the creator's release. Verify before provisioning. More on how we price.

Estimated cost to self-host

Nemotron 3.5 Lightning 30B A3B needs about 19 GB of GPU memory at 4-bit with 32K context. The cheapest rental that fits is 1× RTX 3090 at about $166 a month.

That costs the same as roughly 2B tokens a month on Azure's API. Self-hosting is more expensive below that volume.

Fits on 1 Single node ≤ 8

What it costs to self-host Nemotron 3.5 Lightning 30B A3B at each precision: the memory it needs, the cheapest rental that holds it, that rental's month and the API volume it pays off at
Precision	Memory	Cheapest fit (32K context)(Full 1M context)	Cost /mo	Break-even vs API
4-bit INT4 / FP4	19 GB22 GB	1× RTX 3090	$166$202	2B tokens /mo2B tokens /mo
8-bit FP8 / INT8	32 GB38 GB	2× RTX 3090	$331$331	4B tokens /mo4B tokens /mo
16-bit FP16 / BF16	62 GB68 GB	4× RTX 3090	$662$662	8B tokens /mo8B tokens /mo

Estimates based on median on-demand rates for Nvidia GPUs. Memory is weights plus KV cache for one request, using FP8 KV cache where supported and FP16/BF16 otherwise. Break-even assumes a 5:1 input-to-output ratio. No guarantee of runtime support, usable performance, or that a matching quantized build exists. Pricing methodology.

Similarly priced models

The models nearest Nemotron 3.5 Lightning 30B A3B by blended rate, each at its own cheapest provider.

Models priced nearest Nemotron 3.5 Lightning 30B A3B per 1M tokens, each at its cheapest provider
Model	Blended /1M	Input /1M	Output /1M	Context	Cutoff	vs Nemotron 3.5 Lightning 30B A3B
GPT-OSS-20B OpenAI	$0.0483	$0.03	$0.14	131K	Jun 2024	−44%
Llama 3.2 1B Meta	$0.056	$0.027	$0.201	128K	Dec 2023	−35%
Command R7B Cohere	$0.0563	$0.0375	$0.15	128K		−35%
GPT-OSS-120B OpenAI	$0.0588	$0.039	$0.158	131K	Jun 2024	−32%
Nemotron 3 Nano 30B A3B Nvidia	$0.075	$0.05	$0.20	262K	Jun 2025	−13%
Nemotron 3.5 Lightning 30B A3B This model Nvidia	$0.0867	$0.06	$0.22	1M	Sep 2025
Gemma 4 26B A4B Google Cloud	$0.09	$0.05	$0.29	256K	Jan 2025	+4%
Llama 3.2 3B Meta	$0.0983	$0.051	$0.335	128K	Dec 2023	+13%
Qwen3-30B-A3B FP8 Alibaba Cloud	$0.0983	$0.051	$0.335	33K		+13%
Ministral 3 3B Mistral	$0.10	$0.10	$0.10	256K		+15%
GLM-5.3-Flash Z.AI	$0.1042	$0.075	$0.25	1M		+20%

USD per 1M tokens at each model's cheapest listed provider. Blended is the cost of 10M input plus 2M output tokens, spread over the 12M.

Salad

Access 60,000+ daily active GPUs across 191 countries, from $0.02/hour

Run RTX 50-series and 40-series cards without datacenter pricing

Pick a priority level to balance cost against interruptions

Per-second billing, with no charge for cold starts or image pulls

View Pricing

Capabilities

What Nemotron 3.5 Lightning 30B A3B accepts and can do, as published by Nvidia.

Text Accepts and generates natural-language text.
Reasoning Thinks before it answers, always on or as a switchable mode.
Function calling Connect to external tools, APIs, and systems.

Common questions

What is Nemotron 3.5 Lightning 30B A3B good for?

Agents, coding, math and multi-step tool use in English and code, with a switchable thinking mode. Open weights under the OpenMDW 1.1 license, servable on a single high-memory GPU.

When is Nemotron 3.5 Lightning 30B A3B not a good fit?

Image input: it is text only, and no maximum output is published. NVIDIA meant the full-precision weights mainly for fine-tuning and points serving to its 4-bit build.

What is the cheapest way to run Nemotron 3.5 Lightning 30B A3B?

Hosted, unless you push serious volume. Azure charges $0.06 in / $0.22 out per 1M tokens. The cheapest rental that fits is 1x RTX 3090 at $166 a month, which costs the same as about 2B tokens a month on that API.

How does Nemotron 3.5 Lightning 30B A3B compare with Nemotron 3 Nano 30B A3B?

Same text input. Nemotron 3.5 Lightning 30B A3B costs 20% more per 1M input tokens and 10% more per 1M output tokens than Nemotron 3 Nano 30B A3B, at each model's cheapest listed provider. It came out 8 months later, has a newer knowledge cutoff (September 2025 against June 2025), a larger context (1M against 262K tokens) and fewer parameters (30B with 3B active against 31.6B with 3.6B active) and is licensed under OpenMDW 1.1 against NVIDIA Nemotron Open Model License.

Can I self-host Nemotron 3.5 Lightning 30B A3B?

Yes. The weights are OpenMDW 1.1 licensed. At 4-bit it needs about 19 GB of GPU memory, which starts at roughly $166 a month on the cheapest rental that fits.