Skip to main content

NVIDIA: Nemotron 3 Ultra Coding Benchmark

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...

Context262,144tokens
Max Output16,384tokens
Inputmodality
Price$0.50/1M input

Try NVIDIA: Nemotron 3 Ultra in Kilo Code

Experience this model with the most popular open source coding agent. Free to start, pay only for AI usage. Use in popular IDEs like VS Code, JetBrains, command line, or cloud agents.

3M+

Downloads

500+

models supported

Free

to Start

Access 500+ models including NVIDIA: Nemotron 3 Ultra and many more in Kilo Code

Coding Performance

Coding benchmarks and performance metrics for development tasks

Kilo Bench

% Completion on Terminal Bench 2.0
19.1%
Cost per attempt (USD)
$101.82
Benchmark
Terminal Bench 2.0

Official Kilo eval results. Cost is averaged per complete benchmark attempt.

OpenClaw Benchmarks

PinchBench measures how NVIDIA: Nemotron 3 Ultra performs on real OpenClaw agent tasks: multi-step execution, tool use, recovery, latency, and cost.

Average score

89.9%

#5 of 50 official models

Average time

150m 34s

5 runs · per OpenClaw task

Average cost

Free

Per benchmark run

Category breakdown

Best verified PinchBench v2 run by OpenClaw task family.

Log Analysis95.9% · 5/30 cleared
Productivity95.8% · 4/8 cleared
Memory95.0% · 1/2 cleared
Csv Analysis94.8% · 2/26 cleared

Top task results

Highest-scoring benchmark tasks from the same submission.

Analysis
Access Control Log Anomaly Detection
100.0%
Productivity
Calendar Event Creation
100.0%
Writing
Commit Message Writer
100.0%
Skills
Create Project Structure
100.0%
Coding
Dockerfile Optimization
100.0%
Research
Executive Lookup
100.0%

Autonomous task execution

NVIDIA: Nemotron 3 Ultra shows strong average success across OpenClaw-style benchmark runs, useful for recurring research, browser, and file-based automations.

Tool use and recovery

PinchBench tasks stress multi-step planning, tool calls, and judge-verified completion rather than single prompt coding snippets.

Agent workflow fit

Its deliberate average runtime and cost-aware run cost help set expectations for long-running KiloClaw agents and production workflows.

Agentic benchmarks from the PinchBench Leaderboard

Real-World Usage

Real-world usage statistics from the Kilo Code community

Weekly Token Usage

No ranking data available for this model yet.

Real-world metrics from the Kilo Code Leaderboard

Pricing

Cost per 1 million tokens

Input Tokens
$0.50
per 1M tokens
Output Tokens
$2.20
per 1M tokens

Example Cost

Analyzing a 10,000 line codebase (≈40k input tokens, 10k output tokens) costs approximately $0.0420

Coding Capabilities

Features and parameters relevant to coding tasks

Coding Features

Function Calling
Can call external functions/APIs
Tool Choice
Control over function selection
Structured Outputs
JSON schema validation
Reasoning Tokens
Extended thinking for complex problems

Pricing details from OpenRouter

Technical Details

Architecture and implementation specifications

Specifications
Model ID
nvidia/nemotron-3-ultra-550b-a55b
Created
June 4, 2026
Tokenizer
Other
Input Modalities
Text
Context Window
262,144 tokens
Max Completion Tokens
16,384 tokens
Input Price
$0.50 per 1M tokens
Output Price
$2.20 per 1M tokens
Cache Read Price
$0.10 per 1M tokens
Content Moderation
Disabled

Ready to try NVIDIA: Nemotron 3 Ultra?

Install Kilo Code and start using NVIDIA: Nemotron 3 Ultra for your coding projects today. Choose from 500+ AI models with complete freedom.

  1. Install Kilo Code

    Get the extension from VS Code Marketplace, JetBrains Plugin Repository, or the CLI.

  2. Open the model selector

    Click the model name in the Kilo Code chat panel to open the selector.

  3. Choose your model

    Search or browse to find and select your preferred model.

  4. Start coding

    Use Code, Ask, Debug, or Plan mode — the model is ready immediately.