ModelBench
NVIDIA logo
Models/NVIDIA

Nemotron 3 Ultra 550B A55B

Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy

Source-linked nemotron Open weights Released 2026-06-04
API data Report
Input modalitiesT
Output modalitiesT
Input price$0.5/1M tokens
Output price$2.5/1M tokens
Context window1M
Max output128K
Overview

Capabilities and specifications

✓ Reasoning Yes
✓ Tool calling Yes
✓ Structured output Yes
× Attachments No
× Vision input No
✓ Open weights Yes
Creator
NVIDIA
Model family
nemotron
Knowledge cutoff
Not documented
License
Not documented
Release date
2026-06-04
Model ID
nvidia/nemotron-3-ultra-550b-a55b
Inference availability

Provider pricing and availability

Provider-specific identifiers, limits, and prices per million tokens.

Report a price
Providers offering Nemotron 3 Ultra 550B A55B
ProviderProvider model IDContextInputOutputCache readCapabilitiesSource
UnoRouter nemotron-3-ultra-550b-a55b:free 1M Not listed Not listed - ReasoningTools Docs
Kenari nemotron-3-ultra-550b-a55b 1M Not listed Not listed - ReasoningTools Docs
routing.run nemotron-3-ultra 131.072K $0.1 $0.1 - ReasoningTools Docs
Pioneer nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16 1M $0.5 $2.5 $0.15 ReasoningTools Docs
LLM Gateway nemotron-3-ultra-550b 1.04858M $0.5 $2.5 $0.15 ReasoningTools Docs
Nvidia nvidia/nemotron-3-ultra-550b-a55b 1M $0.5 $2.5 $0.15 ReasoningToolsJSON Docs
Vercel AI Gateway nvidia/nemotron-3-ultra-550b-a55b 1M $0.6 $2.4 $0.12 ReasoningTools Docs
Together AI nvidia/nemotron-3-ultra-550b-a55b 512.3K $0.6 $3.6 $0.2 ReasoningToolsJSON Docs
OpenRouter nvidia/nemotron-3-ultra-550b-a55b 512.288K $0.6 $3.6 $0.2 ReasoningToolsJSON Docs

Capability badges appear only when the provider catalog explicitly lists support.

Serving quality

Performance and reliability

Observed serving measurements stay separate from model specifications and are never estimated from price or model size.

Throughput Not reported

Generation speed; higher is faster.

Latency / TTFT Not reported

Time until the first output token arrives.

Uptime Not reported

Successful requests over a measured period.

Performance varies by inference provider, region, and load. Provider-level measurements appear only when a source reports them.

Published evaluations

Benchmarks

Raw results remain attached to their original source and version.

Benchmark registry
Agentic Index - Unversioned - indexTop 8 reported models
Claude Opus 5
55.3
GPT-5.6 Sol
54.0
Anthropic: Claude Fable 5 (batch)
52.8
Claude Fable 5
52.8
Kimi K3
50.1
GPT-5.6 Terra
47.4
Anthropic: Claude Opus 4.8 (batch)
47.2
Nemotron 3 Ultra 550B A55B
27.4
BrowseComp - Unversioned - accuracyTop 8 reported models
GPT-5.6 Sol
90.4
GPT-5.5 Pro
90.1
GPT-5.4 Pro
89.3
GPT-5.6 Terra
87.5
Claude Sonnet 5
84.7
GPT-5.5
84.4
MiniMax-M3
83.5
Nemotron 3 Ultra 550B A55B
44.4
Coding Index - Unversioned - indexTop 8 reported models
Claude Opus 5
78.0
GPT-5.6 Sol
77.4
GPT-5.6 Terra
76.7
Anthropic: Claude Fable 5 (batch)
76.5
Claude Fable 5
76.5
Kimi K3
76.2
OpenAI: GPT-5.5 (batch)
74.9
Nemotron 3 Ultra 550B A55B
49.3
Design Arena: 3d - Unversioned - eloTop 8 reported models
Kimi K3
1456.0
Claude Opus 5
1384.0
Anthropic: Claude Fable 5 (batch)
1373.0
Claude Fable 5
1373.0
Z.ai: GLM 5.2
1366.0
Z.ai: GLM 5.1
1336.0
Anthropic: Claude Opus 4.6 (batch)
1332.0
Nemotron 3 Ultra 550B A55B
1190.0
Design Arena: asciiart - Unversioned - eloTop 8 reported models
Anthropic: Claude Fable 5 (batch)
1364.0
Claude Fable 5
1364.0
Anthropic: Claude Opus 4.7 (batch)
1324.0
Anthropic: Claude Opus 4.7
1324.0
Muse Spark 1.1
1322.0
Google: Gemini 3.1 Pro Preview (batch)
1309.0
Gemini 3.1 Pro Preview
1309.0
Nemotron 3 Ultra 550B A55B
1113.0
Design Arena: codecategories - Unversioned - eloTop 8 reported models
Kimi K3
1418.0
Claude Opus 5
1358.0
Z.ai: GLM 5.2
1346.0
Anthropic: Claude Fable 5 (batch)
1340.0
Claude Fable 5
1340.0
Anthropic: Claude Opus 4.6 (batch)
1324.0
Anthropic: Claude Opus 4.6
1324.0
Nemotron 3 Ultra 550B A55B
1156.0
Design Arena: dataviz - Unversioned - eloTop 8 reported models
Kimi K3
1382.0
Claude Opus 5
1378.0
Z.ai: GLM 5.1
1366.0
Google: Gemini 3.6 Flash (batch)
1350.0
Gemini 3.6 Flash
1350.0
Anthropic: Claude Fable 5 (batch)
1349.0
Claude Fable 5
1349.0
Nemotron 3 Ultra 550B A55B
1153.0
Design Arena: gamedev - Unversioned - eloTop 8 reported models
Anthropic: Claude Fable 5 (batch)
1393.0
Claude Fable 5
1393.0
Z.ai: GLM 5.2
1358.0
Anthropic: Claude Sonnet 5 (batch)
1352.0
Claude Sonnet 5
1352.0
OpenAI: GPT-5.5 (batch)
1345.0
GPT-5.5
1345.0
Nemotron 3 Ultra 550B A55B
1183.0
Design Arena: svg - Unversioned - eloTop 8 reported models
Anthropic: Claude Fable 5 (batch)
1351.0
Claude Fable 5
1351.0
Google: Gemini 3.1 Pro Preview (batch)
1335.0
Gemini 3.1 Pro Preview
1335.0
Google: Gemini 3.5 Flash (batch)
1296.0
Gemini 3.5 Flash
1296.0
Anthropic: Claude Opus 4.6 (batch)
1276.0
Nemotron 3 Ultra 550B A55B
1127.0
Design Arena: uicomponent - Unversioned - eloTop 8 reported models
Kimi K3
1413.0
Claude Opus 5
1406.0
Anthropic: Claude Fable 5 (batch)
1350.0
Claude Fable 5
1350.0
Anthropic: Claude Opus 4.7 (batch)
1339.0
Anthropic: Claude Opus 4.7
1339.0
Z.ai: GLM 5.2
1334.0
Nemotron 3 Ultra 550B A55B
1178.0
Design Arena: website - Unversioned - eloTop 8 reported models
Kimi K3
1375.0
Claude Opus 5
1343.0
Z.ai: GLM 5.2
1340.0
Anthropic: Claude Fable 5 (batch)
1323.0
Claude Fable 5
1323.0
xAI: Grok 4.5
1323.0
Google: Gemini 3.6 Flash (batch)
1322.0
Nemotron 3 Ultra 550B A55B
1129.0
GDPval - Unversioned - wins or tiesTop 6 reported models
GPT-5.5
84.9
GPT-5.4
83.0
GPT-5.5 Pro
82.3
GPT-5.4 Pro
82.0
Nemotron 3 Ultra 550B A55B
46.7
Step 3.7 Flash
45.8
GPQA - Unversioned - accuracyTop 1 reported models
Nemotron 3 Ultra 550B A55B
87.0
Humanity's Last Exam - Unversioned - accuracyTop 8 reported models
Claude Fable 5
64.5
Muse Spark 1.1
62.1
GPT-5.4 Pro
58.7
Claude Opus 4.8
57.9
GPT-5.5 Pro
57.2
Claude Opus 4.7
54.7
GPT-5.5
52.2
Nemotron 3 Ultra 550B A55B
37.4
IFBench - Unversioned - accuracyTop 2 reported models
Nemotron 3 Ultra 550B A55B
81.7
Grok 4.3
81.0
Intelligence Index - Unversioned - indexTop 8 reported models
Claude Opus 5
60.7
Anthropic: Claude Fable 5 (batch)
59.9
Claude Fable 5
59.9
GPT-5.6 Sol
58.9
Kimi K3
57.1
Anthropic: Claude Opus 4.8 (batch)
55.7
Anthropic: Claude Opus 4.8
55.7
Nemotron 3 Ultra 550B A55B
37.8
LiveCodeBench - v6 - pass@1Top 1 reported models
Nemotron 3 Ultra 550B A55B
89.0
MMLU-Pro - Unversioned - accuracyTop 1 reported models
Nemotron 3 Ultra 550B A55B
86.8
SWE-Bench Multilingual - Unversioned - resolve rateTop 5 reported models
Claude Sonnet 5
78.3
Qwen3.7 Max
78.3
Grok 4.5
78.0
LongCat-2.0
77.3
Nemotron 3 Ultra 550B A55B
67.7
SWE-Bench Verified - Unversioned - resolvedTop 8 reported models
Claude Fable 5
95.0
Claude Opus 4.8
88.6
Claude Sonnet 5
85.2
DeepSeek V4 Pro
80.6
Qwen3.7 Max
80.4
Kimi K2.6
80.2
DeepSeek V4 Flash
79.0
Nemotron 3 Ultra 550B A55B
70.7
Terminal-Bench - 2.1 - success rateTop 8 reported models
GPT-5.6 Sol
88.8
Claude Fable 5
88.0
GPT-5.6 Terra
87.4
GPT-5.6 Luna
84.7
Grok 4.5
83.3
Muse Spark 1.1
80.0
LongCat-2.0
70.8
Nemotron 3 Ultra 550B A55B
56.4
Catalog activity

Recent record changes

Detected changes from successful source imports.

Full change log
Context Length512288 → 1000000
Price Completion3.5999999999999996 → 2.5
Price Prompt0.6 → 0.5
Context Length1000000 → 512288
Price Completion2.5 → 3.5999999999999996
Price Prompt0.5 → 0.6
Context Length512288 → 1000000
Price Completion3.5999999999999996 → 2.5
Public API

Use this record

Fetch the complete source-linked model record without an API key.

API documentation
Endpoint
GET https://www.modelbench.lol/api/v1/models/nvidia/nemotron-3-ultra-550b-a55b
curl
curl "https://www.modelbench.lol/api/v1/models/nvidia/nemotron-3-ultra-550b-a55b"