Nemotron 3 Ultra 550B A55B

NVIDIA Nemotron 3 Ultra is an open frontier-reasoning and orchestration model from NVIDIA, with 55B active parameters out of 550B total (MoE). Built on a hybrid Transformer-Mamba mixture-of-experts architecture, it supports text input and output with a context window of up to 1M tokens. It is suited for long-running agentic workflows, including agent orchestration, coding agents, deep research, and complex enterprise tasks.

It is particularly strong at multi-step reasoning and planning, with high-throughput inference designed for high-volume agent pipelines. It is part of the NVIDIA Nemotron family of open models for agentic AI.

nvidia/nemotron-3-ultra-550b-a55b

Organization: Nvidia
Family: nemotron
Providers: 21
Context: 1,000,000
Output limit: 128,000
Knowledge: —
Release: 2026-06-04
Updated: 2026-06-04
Weights: Open
Input: text
Output: text
Capabilities: Tools, Reasoning, Temperature

Providers

Compare every cataloged offer for this model, including its provider, route, token limits, price, and capabilities.

27 offers · USD per million tokens

Provider	Model ID	Context	Output	Input / output · 1M	Reasoning	Tools	Structured	Details
Baseten	nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B	202,800	202,800	$0.60 / $2.40	Yes	Yes	Yes	View
Bothub	nemotron-3-ultra-550b-a55b:free	1,000,000	128,000	$0.00 / $0.00	Yes	Yes	—	View
CoreWeave	nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B	262,144	262,144	$0.50 / $2.15	Yes	Yes	Yes	View
DevPass (LLM Gateway)	nemotron-3-ultra-550b	1,048,576	128,000	$0.50 / $2.20	Yes	Yes	—	View
Eden AI	deepinfra/nemotron-3-ultra-550b-a55b	262,144	128,000	$0.50 / $2.20	Yes	Yes	Yes	View
Eden AI	nebius/nvidia/Nemotron-3-Ultra-550b-a55b	1,048,576	128,000	$1.00 / $3.00	Yes	Yes	Yes	View
Fireworks AI	accounts/fireworks/models/nemotron-3-ultra-nvfp4	262,144	128,000	$0.60 / $2.40	Yes	Yes	—	View
Kenari	nemotron-3-ultra-550b-a55b	1,000,000	128,000	$0.00 / $0.00	Yes	Yes	—	View
Kilo Gateway	nvidia/nemotron-3-ultra-550b-a55b:free	1,000,000	65,536	$0.00 / $0.00	Yes	Yes	No	View
Kilo Gateway	nvidia/nemotron-3-ultra-550b-a55b	262,144	16,384	$0.50 / $2.20	Yes	Yes	Yes	View
NanoGPT	nvidia/nemotron-3-ultra-550b-a55b	1,000,000	65,536	$0.50 / $2.50	Yes	Yes	No	View
NanoGPT	nvidia/nemotron-3-ultra-550b-a55b:thinking	1,000,000	65,536	$0.50 / $2.50	Yes	Yes	No	View
Nebius Token Factory	nvidia/Nemotron-3-Ultra-550b-a55b	1,048,576	1,048,576	$1.00 / $3.00	Yes	Yes	—	View
Nvidia	nvidia/nemotron-3-ultra-550b-a55b	1,000,000	65,536	$0.50 / $2.50	Yes	Yes	Yes	View
Ollama Cloud	nemotron-3-ultra	262,144	128,000	$0.10 / $3.00	Yes	Yes	—	View
OpenCode Zen	nemotron-3-ultra-free	1,000,000	128,000	$0.00 / $0.00	Yes	Yes	—	View
BaseTenvia OpenRouter	nvidia/nemotron-3-ultra-550b-a55b fp4	202,800	182,520	$0.60 / $2.40	Yes	Yes	No	View
DeepInfravia OpenRouter	nvidia/nemotron-3-ultra-550b-a55b fp4	262,144	16,384	$0.50 / $2.20	Yes	Yes	Yes	View
Nvidiavia OpenRouter	nvidia/nemotron-3-ultra-550b-a55b:free unknown	1,000,000	65,536	$0.00 / $0.00	Yes	Yes	No	View
Venicevia OpenRouter	nvidia/nemotron-3-ultra-550b-a55b fp8	256,000	32,768	$0.625 / $3.125	Yes	Yes	No	View
Pioneer	nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16	1,000,000	65,000	$0.50 / $2.50	Yes	Yes	—	View
Requesty	nemotron-3-ultra-550b-a55b	1,048,576	65,536	$0.00 / $0.00	Yes	Yes	No	View
routing.run	nemotron-3-ultra	131,072	32,000	$0.10 / $0.10	Yes	Yes	—	View
Together AI	nvidia/nemotron-3-ultra-550b-a55b	512,300	512,300	$0.60 / $3.60	Yes	Yes	Yes	View
UnoRouter	nemotron-3-ultra-550b-a55b:free	1,000,000	128,000	$0.00 / $0.00	Yes	Yes	—	View
Venice AI	nvidia-nemotron-3-ultra-550b-a55b	256,000	32,768	$0.625 / $3.125	Yes	Yes	Yes	View
Vercel AI Gateway	nvidia/nemotron-3-ultra-550b-a55b	1,000,000	65,000	$0.60 / $2.40	Yes	Yes	—	View

Prices and limits apply to each offer. A dash means the value is not provided. OpenRouter offers show the hosting provider and routing channel separately.

Pricing

Compare token rates and cache charges across available routes and hosting configurations.

USD per million tokens

Input from$0.00Per 1M input tokens

Output from$0.00Per 1M output tokens

Available offers27Across 21 providers

Lowest listed input and output prices across the offers above; they may belong to different providers. These are base rates, before conditional pricing, discounts or additional fees.

Endpoint price comparisonLatest observed snapshot · 4 lowest-priced configurations

InputOutput

Nvidianvidia

$0.00$0.00

DeepInfradeepinfra/fp4

$0.50$2.20

BaseTenbaseten/fp4

$0.60$2.40

Venicevenice/fp8

$0.625$3.125

nvidia/nemotron-3-ultra-550b-a55bObserved Oct 4, 2026, 11:31 AM UTC

Input: $0.50
Output: $2.20
Cache read: $0.10

nvidia/nemotron-3-ultra-550b-a55b:freeObserved Oct 4, 2026, 11:31 AM UTC

Input: $0.00
Output: $0.00

Hosting prices through OpenRouter All endpoints and cache rates

Endpoint	Input / 1M	Output / 1M	Cache read / 1M	Cache write / 1M
BaseTenbaseten/fp4	$0.60	$2.40	$0.12	—
DeepInfradeepinfra/fp4	$0.50	$2.20	$0.10	—
Venicevenice/fp8	$0.625	$3.125	$0.1875	—
Nvidianvidia	$0.00	$0.00	—	—

Each row is a hosting configuration. Open an offer in Providers for its full pricing conditions and additional fees. Endpoint observations may differ from the routing catalog's base price.

Performance

Review reported latency and output-throughput percentiles from the latest observed rolling window.

Latest 30-minute observation

Time to first token · milliseconds

Latency distributionRolling 30-minute snapshot · 3 endpoints

P50P75P90P99

DeepInfradeepinfra/fp4

1,462.5–11,137.72 ms

Venicevenice/fp8

1,465.5–5,264.6 ms

Nvidianvidia

1,921–26,643.89 ms

All latency percentiles 3 endpoints

Endpoint	P50	P75	P90	P99
DeepInfradeepinfra/fp4	1,462.5 ms	2,147.25 ms	3,111.5 ms	11,137.72 ms
Venicevenice/fp8	1,465.5 ms	1,817.75 ms	2,650.5 ms	5,264.6 ms
Nvidianvidia	1,921 ms	3,098 ms	5,613 ms	26,643.89 ms

Output throughput · tokens per second

Throughput distributionRolling 30-minute snapshot · 3 endpoints

P50P75P90P99

DeepInfradeepinfra/fp4

79–173.5 t/s

Venicevenice/fp8

97–174.44 t/s

Nvidianvidia

29–75 t/s

All throughput percentiles 3 endpoints

Endpoint	P50	P75	P90	P99
DeepInfradeepinfra/fp4	79 t/s	112 t/s	134 t/s	173.5 t/s
Venicevenice/fp8	97 t/s	118 t/s	134 t/s	174.44 t/s
Nvidianvidia	29 t/s	40 t/s	52 t/s	75 t/s

Percentiles describe the observed request distribution. These are snapshots of a rolling window, not a historical time series.

Uptime

Compare successful request rates across the latest 5-minute, 30-minute, and 24-hour observations.

Successful requests by window

Availability by endpointCurrent overlapping windows · 4 endpoints

Endpoint5 minutes30 minutes24 hours

BaseTenbaseten/fp4Observed Oct 4, 2026, 11:31 AM UTC

—

100%

—

DeepInfradeepinfra/fp4Observed Oct 4, 2026, 11:31 AM UTC

100%

99.95%

Venicevenice/fp8Observed Oct 4, 2026, 11:31 AM UTC

—

100%

99.89%

NvidianvidiaObserved Oct 4, 2026, 11:31 AM UTC

99.02%

98.29%

98.15%

Reported by OpenRouter at the time of collection; rate-limited requests are excluded. A dash means no measurement was reported. This is not a live availability check.

Benchmarks

Review attributed evaluation results on each source's original scale.

External evaluations

nvidia/nemotron-3-ultra-550b-a55bObserved Oct 4, 2026, 11:31 AM UTC

Benchmark profileArtificial Analysis · shared source scale from 0 to 50

Coding

49.3

Agentic

20.1

Intelligence

22.9

Design Arena	Category	Elo	Win rate	Rank
models	3d	1,147	39.3%	65
models	asciiart	1,103	37.5%	59
models	codecategories	1,150	36.2%	77
models	dataviz	1,149	37.5%	77
models	gamedev	1,152	36.9%	68
models	svg	1,070	32.9%	68
models	uicomponent	1,147	36.6%	74
models	website	1,146	34.9%	85

nvidia/nemotron-3-ultra-550b-a55b:freeObserved Oct 4, 2026, 11:31 AM UTC

Benchmark profileArtificial Analysis · shared source scale from 0 to 50

Coding

49.3

Agentic

20.1

Intelligence

22.9

Design Arena	Category	Elo	Win rate	Rank
models	3d	1,147	39.3%	65
models	asciiart	1,103	37.5%	59
models	codecategories	1,150	36.2%	77
models	dataviz	1,149	37.5%	77
models	gamedev	1,152	36.9%	68
models	svg	1,070	32.9%	68
models	uicomponent	1,147	36.6%	74
models	website	1,146	34.9%	85

Sources: Artificial Analysis and Design Arena, via OpenRouter. Scores retain their original scales and are not OpenAI Suite ratings.

Detailed evaluations

Evaluation scoresAccuracy or item F1 · higher is better

tau bench verified airlineNVIDIA: Nemotron 3 Ultra

72%

gpqa diamondNVIDIA: Nemotron 3 Ultra

81.202%

Evaluation	Configuration	Source	Results	Details
Composite indices	Nemotron 3 Ultra 550B A55B (Reasoning)nvidia/nemotron-3-ultra-550b-a55b-20260604	Artificial Analysis	Intelligence 22.9 · Coding 49.3 · Agentic 20.1	Details
models · dataviz	Nemotron 3 Ultranvidia/nemotron-3-ultra-550b-a55b-20260604	Design Arena	1,149 Elo	Details
tau bench verified airline	NVIDIA: Nemotron 3 Ultranvidia/nemotron-3-ultra-550b-a55b-20260604	OpenRouter	72%	Details
models · svg	Nemotron 3 Ultranvidia/nemotron-3-ultra-550b-a55b-20260604	Design Arena	1,070 Elo	Details
models · 3d	Nemotron 3 Ultranvidia/nemotron-3-ultra-550b-a55b-20260604	Design Arena	1,147 Elo	Details
models · codecategories	Nemotron 3 Ultranvidia/nemotron-3-ultra-550b-a55b-20260604	Design Arena	1,150 Elo	Details
models · website	Nemotron 3 Ultranvidia/nemotron-3-ultra-550b-a55b-20260604	Design Arena	1,146 Elo	Details
models · uicomponent	Nemotron 3 Ultranvidia/nemotron-3-ultra-550b-a55b-20260604	Design Arena	1,147 Elo	Details
models · gamedev	Nemotron 3 Ultranvidia/nemotron-3-ultra-550b-a55b-20260604	Design Arena	1,152 Elo	Details
gpqa diamond	NVIDIA: Nemotron 3 Ultranvidia/nemotron-3-ultra-550b-a55b-20260604	OpenRouter	81.202%	Details
models · asciiart	Nemotron 3 Ultranvidia/nemotron-3-ultra-550b-a55b-20260604	Design Arena	1,103 Elo	Details

Sources: Artificial Analysis, Design Arena and OpenRouter. Evaluation configurations and scales differ; results are not interchangeable. Updated Oct 4, 2026.

Apps & session costs

Compare observed median session costs for applications using this model, grouped by conversation length.

30-day sample · through Sep 27, 2026

Application	Model route	Turns	Median cost / session	Details
Hermes Agent	nvidia/nemotron-3-ultra-550b-a55b-20260604	1 turn	$0.007828	Details
Hermes Agent	nvidia/nemotron-3-ultra-550b-a55b-20260604	10–49 turns	$0.316059	Details
Hermes Agent	nvidia/nemotron-3-ultra-550b-a55b-20260604	2–9 turns	$0.021615	Details

Source: OpenRouter, as of Oct 2, 2026. Published under CC BY 4.0. These are observed costs per session, not token prices or a forecast for your workload. Apps are compared separately.

Activity

Daily prompt and completion tokens reported for this model's routes in OpenRouter's top 50. Missing days have no published value.

Last 90 complete UTC days

Daily token usage23,423,612,244 reported tokens · 1 days with observations

23B0

2026-07-09
23,423,612,244 tokens

Jul 6, 2026Oct 3, 2026

Daily observations · 1 days

Date (UTC)	Reported tokens	Reported routes
Jul 9, 2026	23,423,612,244	1

Source: OpenRouter (openrouter.ai/rankings), as of Oct 4, 2026. CC BY 4.0. Totals include only individually published routes. Tokenizers differ by provider; the aggregated “other” category is never assigned to a model.

Technical details & API

Explore specifications, supported parameters, reasoning controls, and API identifiers for each OpenRouter route.

Current route specifications

nvidia/nemotron-3-ultra-550b-a55bObserved Oct 4, 2026, 11:31 AM UTC

Context: 262,144
Maximum output: 16,384
Tokenizer: Other
Input: text
Output: text
Moderation: No
Added to OpenRouter: 2026-06-04

Base URLhttps://openrouter.ai/api/v1

Model IDnvidia/nemotron-3-ultra-550b-a55b

Version IDnvidia/nemotron-3-ultra-550b-a55b-20260604

Weights IDnvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16

Reasoning

Required: No
Enabled by default: Yes
Default effort: high
Supported efforts: high, medium

Supported parameters

frequency_penalty
include_reasoning
logit_bias
max_tokens
min_p
presence_penalty
reasoning
reasoning_effort
repetition_penalty
response_format
seed
stop
structured_outputs
temperature
tool_choice
tools
top_k
top_p

Default parameters

top p: 0.95
temperature: 1

nvidia/nemotron-3-ultra-550b-a55b:freeObserved Oct 4, 2026, 11:31 AM UTC

Context: 1,000,000
Maximum output: 65,536
Tokenizer: Other
Input: text
Output: text
Moderation: No
Added to OpenRouter: 2026-06-04

Base URLhttps://openrouter.ai/api/v1

Model IDnvidia/nemotron-3-ultra-550b-a55b:free

Version IDnvidia/nemotron-3-ultra-550b-a55b-20260604

Weights IDnvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16

Reasoning

Required: No
Enabled by default: Yes
Default effort: high
Supported efforts: high, medium

Supported parameters

include_reasoning
max_tokens
reasoning
reasoning_effort
seed
temperature
tool_choice
tools
top_p

Default parameters

top p: 0.95
temperature: 1

Frequently asked questions

Answers based on the model metadata and serving offers currently in the catalog.

How much does Nemotron 3 Ultra 550B A55B cost?

Listed rates start at $0.00 per million input tokens and $0.00 per million output tokens. Prices vary by provider and configuration; see Pricing for details.

Which providers offer Nemotron 3 Ultra 550B A55B?

There are 27 cataloged offers across 21 providers. Compare model IDs, prices and limits in Providers.

What is the context window?

The cataloged model context is 1,000,000 tokens. Each serving endpoint may apply a different limit.

Does it support tools and structured output?

Tool calling: Yes. Structured output: Not reported. Support can vary by endpoint.

Catalog history · 2 recorded revisions

Metadata observations since this model was first cataloged. These are not model release versions.

Sep 14, 2026 · e7b0d6404885
Sep 12, 2026 · e07af669d579