Explore Reasoning Models | Try NVIDIA NIM APIs

Skip to main content

Explore Reasoning Models | Try NVIDIA NIM APIs

Deploy Models Now with NVIDIA NIM

Optimized inference for the world’s leading models

Free serverless APIs for development

Accelerated by DGX Cloud

Self-Host on your GPU infrastructure

Continuous vulnerability fixes

Discover Models Skills Blueprints GPUs Docs Forums

workstations

Run on RTX
Run on Spark
Run on Station

models

Reasoning
Vision
Visual Design
Retrieval
Speech
Biology
Simulation
Climate & Weather
Safety & Moderation

industries

Automotive
Financial Services
Gaming
Healthcare
Industrial
Robotics

Your Privacy Choices

Copyright © 2026 NVIDIA Corporation

Reasoning

Reasoning

Explore

Workstations

Run on RTX Run on Spark Run on Station

Models

Reasoning Vision Visual Design Retrieval Speech Biology Simulation Climate & Weather Safety & Moderation

Industries

Automotive Financial Services Gaming Healthcare Industrial Robotics

Developer Favorites

The top large language models for your enterprise AI

DownloadableFree Endpoint

kimi-k3

~2.8T hybrid KDA+MLA multimodal MoE for long-horizon coding, agentic tool use, and image understanding.

Mixture-of-Experts

Multimodal
Reasoning

Last updated on August 27, 2026

DownloadableFree Endpoint

deepseek-v4-pro-0813

DeepSeek V4 scales to 262K-token context windows with efficient MoE architecture for coding tasks.

Moe

agentic
coding
reasoning

Last updated on August 26, 2026

DownloadableFree Endpoint

nemotron-3.5-lightning-30b-a3b

Fastest 30B A3B MoE model with leading domain accuracy for specialized agentic tasks

Customization

Long-running agents
Open
Text-to-Text

Last updated on August 11, 2026

DownloadableFree Endpoint

nemotron-3-ultra-550b-a55b

Open, efficient hybrid Mamba-Transformer MoE with 1M context, excelling in agentic reasoning, coding, planning, tool calling, and more

Agent

Frontier
Long Context
MoE
Reasoning

52M API calls in the last 30 days

Last updated on June 4, 2026

DownloadableFree Endpoint

cosmos3-nano-reasoner

Vision language model that excels in understanding the physical world using structured reasoning on videos or images.

Physical AI

autonomous vehicles
industrial
reasoning
robotics
smart cities
video understanding
vision language model

2K API calls in the last 30 days

Last updated on June 1, 2026

DownloadableFree Endpoint

nemotron-3-nano-omni-30b-a3b-reasoning

Nemotron 3 Nano Omni is an omni-modal reasoning model that understands images, video, speech, text.

OCR

Omni
VLM
Video

8M API calls in the last 30 days

Last updated on April 28, 2026

DownloadableFree Endpoint

gemma-4-31b-it

Dense 31B model delivering frontier reasoning for coding, agentic workflows, and fine-tuning.

agentic

coding
reasoning
text-to-text

6M API calls in the last 30 days

Last updated on April 2, 2026

DownloadableFree Endpoint

gpt-oss-20b

Smaller Mixture of Experts (MoE) text-only LLM for efficient AI reasoning and math

chat

math
reasoning
text-to-text

19M API calls in the last 30 days

Last updated on August 5, 2025

Fresh Off the Press

The latest innovations in intelligence models

DownloadableFree Endpoint

kimi-k3

~2.8T hybrid KDA+MLA multimodal MoE for long-horizon coding, agentic tool use, and image understanding.

DownloadableFree Endpoint

deepseek-v4-pro-0813

DeepSeek V4 scales to 262K-token context windows with efficient MoE architecture for coding tasks.

Downloadable

wan2.2-animate-2-14b

Wan2.2-Animate-2 is a novel end-to-end character animation framework

Deprecation in 4dFree Endpoint

deepseek-v4-flash-0731

284B MoE (13B active) model ideal for long-context workloads optimized for coding, chat, and agentic workflows

DownloadableFree Endpoint

nemotron-3.5-lightning-30b-a3b

Fastest 30B A3B MoE model with leading domain accuracy for specialized agentic tasks

DownloadableFree Endpoint

muse-glimmer-30b

Muse Glimmer 30B is a multimodal reasoning model accepting text and images, with native tool-calling and separate reasoning output.

Free Endpoint

riva-translate-4b-instruct-v2

Translation model in 37 languages with few-shots example prompts capability.

Free Endpoint

ising-calibration-1.5-31b

NVIDIA-Ising-Calibration-1.5 is a dense multimodal vision-language model built on Gemma 4 31B. It analyzes quantum computing calibration experiment plots and generates structured technical text.

Downloadable

Video Super Resolution NIM

Upscale encoded or ST 2110 video to higher resolutions with NVIDIA Video Super Resolution.

Free Endpoint

nemotron-3-embed-1b

1B embedding model for semantic search, retrieval, and RAG applications.

Mixture-of-Experts

Multimodal
Reasoning

Last updated on August 27, 2026

Moe

agentic
coding
reasoning

Last updated on August 26, 2026

character animation

video editing

Last updated on August 19, 2026

Hybrid Attention

Long Context
MoE
Reasoning

Last updated on August 19, 2026

Customization

Long-running agents
Open
Text-to-Text

Last updated on August 11, 2026

Chat

Image-to-Text
Large Language Models
Multimodal
Reasoning
Text-to-Text

Last updated on August 10, 2026

neural machine translation

nvidia nim

Last updated on July 27, 2026

Calibration

NVIDIA NIM
Quantum Computing
Vision Language Model

Last updated on July 23, 2026

broadcast

nvidia ai for media
streaming
video super resolution
video upscaling

Last updated on July 22, 2026

Agentic Retrieval

Code Retrieval
Nemotron Retriever

Last updated on July 16, 2026