NVIDIA modelsimage generation api

nvidia/nemotron-3-nano-omni/audio

Name: Nemotron 3 Nano Omni
Author: NVIDIA

Nemotron 3 Nano Omni Audio by NVIDIA - advanced AI model for audio-to-text. Delivers high-quality results with fast inference, suitable for both creative and production workflows.

Input

Prompt(string)

Text prompt to send to the model. English only.

Audio URL(string)

URL of the audio clip to reason about. wav or mp3, max 1 hour.

Top P(number)

Nucleus sampling probability mass. Range: 0 to 1.

Temperature(number)

Sampling temperature. Lower is more deterministic. Range: 0 to 2.

System Prompt(string)

Optional system prompt to steer the model. Reasoning behavior is controlled by the separate `reasoning_mode` field.

Reasoning Mode(string)

Whether the model should emit an explicit reasoning trace. `no_think` returns a direct answer; `think` returns chain-of-thought followed by the final answer. Allowed values: think, no_think.

Max Tokens(integer)

120000

Maximum number of tokens to generate. Range: 1 to 20000.

Idle

Example output — click Run to generate your own

Production endpoint

Send your first request

OpenAI-compatible endpoint with unified authentication and usage tracking.

POSThttps://api.sandbase.ai/v1/run

Model IDnvidia/nemotron-3-nano-omni/audio

Input Schema

7 parameters · 1 required · 6 optional

Parameter	Type	Required	Description
`prompt`	string	Required	Text prompt to send to the model. English only. · Min length: 1
`audio`	string	Optional	URL of the audio clip to reason about. wav or mp3, max 1 hour.
`top_p`	number	Optional	Nucleus sampling probability mass. · Min: 0 · Max: 1 · Default: 0.95
`max_tokens`	integer	Optional	Maximum number of tokens to generate. · Min: 1 · Max: 20000 · Default: 1024
`temperature`	number	Optional	Sampling temperature. Lower is more deterministic. · Min: 0 · Max: 2 · Default: 0.7
`system_prompt`	string	Optional	Optional system prompt to steer the model. Reasoning behavior is controlled by the separate `reasoning_mode` field.
`reasoning_mode`	string	Optional	Whether the model should emit an explicit reasoning trace. `no_think` returns a direct answer; `think` returns chain-of-thought followed by the final answer. · Options: think, no_think · Default: "no_think" thinkno_think

Output Schema

Field	Type	Description
`id`	string	Unique identifier for the generation task
`status`	string	Task status: pending, running, completed, failed, timeout
`model`	string	Model used for the generation
`outputs`	array	Array of output items
`outputs[].url`	string	URL of the generated artifact
`outputs[].content_type`	string	MIME type (e.g. image/png, video/mp4)
`error`	object \| null	Error details if failed, null on success
`error.type`	string	Machine-readable error type code
`error.message`	string	Human-readable error description

Async Workflow

This model uses asynchronous execution. Submit a request and poll for the result.

Submit — POST to /v1/run, receive an id
Poll — GET /v1/run/{id} until status is completed, failed, or timeout
Retrieve — Read outputs from the completed response

Code Examples

Ready-to-run snippets

# Step 1: Submit
curl -X POST https://api.sandbase.ai/v1/run \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -d '{
  "model": "nvidia/nemotron-3-nano-omni/audio",
  "audio": "https://storage.googleapis.com/falserverless/example_inputs/nemotron-3-nano-omni/audio_in.mp3",
  "top_p": 0.95,
  "prompt": "Summarize the key capabilities of a multimodal agent.",
  "max_tokens": 1024,
  "temperature": 0.7,
  "system_prompt": "You are a concise enterprise assistant.",
  "reasoning_mode": "no_think"
}'

# Step 2: Poll result (replace <id>)
curl https://api.sandbase.ai/v1/run/<id> \
  -H "Authorization: Bearer YOUR_API_KEY"

← Back to Nemotron 3 Nano Omni

API README

Nemotron 3 Nano Omni Audio

Nemotron 3 Nano Omni Audio is built for multimodal reasoning, with this exact route focused on spoken-audio understanding. Nemotron 3 Nano Omni Audio by NVIDIA - advanced AI model for audio-to-text. Delivers high-quality results with fast inference, suitable for both creative and production workflows. It is designed for production workflows that need a clearly scoped media operation instead of a generic assistant response, preserving the supplied direction while returning an artifact ready for review, iteration, or downstream assembly.

The request surface keeps creative direction and delivery choices explicit through top_p, prompt, audio_url, sync_mode, max_tokens, temperature. Teams can lock source material, describe the intended result, choose supported output controls, and rerun the same request structure across variations. This makes experiments easier to compare and larger pipelines easier to operate without hiding important decisions inside an opaque preset.

Highlights

Spoken-Audio Understanding. The nvidia/nemotron-3-nano-omni/audio route is dedicated to spoken-audio understanding, keeping inputs and outputs matched to this operation.
Instruction-Following Analysis. Within nvidia/nemotron-3-nano-omni/audio, documented prompt and media controls direct instruction-following analysis while keeping the brief close to the result.
Structured Natural-Language Answers. The nvidia/nemotron-3-nano-omni/audio result provides structured natural-language answers in a machine-consumable form for review and downstream production.
Route-specific control. For nvidia/nemotron-3-nano-omni/audio, documented controls include top_p, prompt, audio_url, sync_mode, max_tokens, temperature, defining the supported variation space for this exact route.

Pricing

Billing status	Price
Current quote	Check the console before submitting.

When to Use

Scenario	Why it fits
Choose this route	When the deliverable is specifically spoken-audio understanding, not a neighboring generation or editing task.
Production iteration	When creative teams need repeatable instruction-following analysis with an explicit request payload.
Automated pipelines	When structured natural-language answers must flow into review, storage, compositing, or publishing.
Controlled variations	When a stable brief should be rerun while changing only documented inputs.
Prototype to scale	When the same `nvidia/nemotron-3-nano-omni/audio` contract should serve a single experiment and a larger batch later.

Prompt Guide

Lead with the subject or source, then state the transformation, important spatial or temporal details, style, and finish. Supply only documented fields, keep media URLs reachable, and vary one control at a time when comparing outputs.

{
  "prompt": "A directed creative request with subject, setting, style, motion, and finish."
}

Technical Specs

Property	Value
Model ID	`nvidia/nemotron-3-nano-omni/audio`
Required inputs	`prompt`
`top_p`	number; default: 0.95
`prompt`	string
`audio_url`	string
`sync_mode`	boolean; default: false
`max_tokens`	integer; default: 1024
`temperature`	number; default: 0.7
`system_prompt`	string
`reasoning_mode`	string; options: think, no_think; default: no_think
`safety_tolerance`	string; default: 6

Related Models

nvidia/nemotron-3-nano-omni/video
nvidia/nemotron-3-nano-omni/vision