NVIDIA
Explore Models Blueprints GPUs Docs

Deploy Models Now with NVIDIA NIM

Optimized inference for the world’s leading models
Free serverless APIs for developmentAccelerated by DGX Cloud
Self-Host on your GPU infrastructure
Continuous vulnerability fixes
DiscoverModelsBlueprintsGPUsDocsForums
models
ReasoningVisionVisual DesignRetrievalSpeechBiologySimulationClimate & WeatherSafety & ModerationRun on RTX
industries
AutomotiveGamingHealthcareIndustrialRobotics

Reasoning

Terms of Use

|

Privacy Policy

|

Manage My Privacy

|

Contact

Copyright © 2025 NVIDIA Corporation

Developer Favorites

The top large language models for your enterprise AI

Run Anywhere

openaigpt-oss-120b

Mixture of Experts (MoE) reasoning LLM (text-only) designed to fit within 80GB GPU.

mathchatreasoningtext-to-text
Run Anywhere

openaigpt-oss-20b

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

mathchatreasoningtext-to-text
Run Anywhere

nvidiallama-3.3-nemotron-super-49b-v1.5

High efficiency model with leading accuracy for reasoning, tool calling, chat, and instruction following.

advanced reasoningfunction callinginstruction followingmath
PREVIEW

moonshotaikimi-k2-instruct

State-of-the-art open mixture-of-experts model with strong reasoning, coding, and agentic capabilities

advanced reasoningagenticcodingchat

Combine Vision and Language Intelligence

Multimodal reasoning models

PREVIEW

microsoftkosmos-2

Groundbreaking multimodal model designed to understand and reason about visual elements in images.

image understandingmultimodalvisual question answeringcomputer visioncvimageimage-to-textvideovlm
PREVIEW

nvidianeva-22b

Multi-modal vision-language model that understands text/images and generates informative responses

non-commercial use onlyvision assistantvisual question answeringcomputer visioncvimageimage-to-textvideovlm
PREVIEW

googledeplot

Translate images of plots into tables with one-shot visual language understanding.

multimodaldata ingestionextractionnemo retrieverimage-to-text

Fresh Off the Press

The latest innovations in intelligence models

Run Anywhere

microsoftTRELLIS

MSFT TRELLIS is a 3D AI model that generates high-quality 3D assets from text or image inputs.

run-on-rtximage-to-3dtext-to-3d
PREVIEW

qwenqwen3-coder-480b-a35b-instruct

Excels in agentic coding and browser use and supports 256K context, delivering top results.

agentic codingbrowser uselong contextmoe
PREVIEW

deepseek-aideepseek-v3.1

DeepSeek V3.1 Instruct is a hybrid AI model with fast reasoning, 128K context, and strong tool use.

reasoningtext-to-text
Run Anywhere

openaigpt-oss-120b

Mixture of Experts (MoE) reasoning LLM (text-only) designed to fit within 80GB GPU.

mathchatreasoningtext-to-text
Run Anywhere

openaigpt-oss-20b

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

mathchatreasoningtext-to-text
Run Anywhere

nvidiallama-3.3-nemotron-super-49b-v1.5

High efficiency model with leading accuracy for reasoning, tool calling, chat, and instruction following.

advanced reasoningfunction callinginstruction followingmath
PREVIEW

moonshotaikimi-k2-instruct

State-of-the-art open mixture-of-experts model with strong reasoning, coding, and agentic capabilities

advanced reasoningagenticcodingchat
PREVIEW

metallama-4-maverick-17b-128e-instruct

A general purpose multimodal, multilingual 128 MoE model with 17B parameters.

language generationimage-to-textvision assistantvisual question answering
Run Anywhere

metallama-4-scout-17b-16e-instruct

A multimodal, multilingual 16 MoE model with 17B parameters.

language generationimage-to-textvision assistantvisual question answering
PREVIEW

deepseek-aideepseek-r1-0528

Updated version of DeepSeek-R1 with enhanced reasoning, coding, math, and reduced hallucination.

advanced reasoningcodingmath