# NVIDIA Nemotron 3.5 Super Vision

> Ask NVIDIA Nemotron 3.5 Super questions about an image: describe scenes, read text and charts, and reason over what it sees.


## Overview

- **Endpoint**: `https://fal.run/nvidia/nemotron-3-5-super/vision`
- **Model ID**: `nvidia/nemotron-3-5-super/vision`
- **Category**: image-to-text
- **Kind**: inference
**Description**: NVIDIA Nemotron 3.5 Super vision route, served by fal. Inputs: image_url (RGB JPEG or PNG) and a prompt about the image, plus an optional system prompt. Shared options: system_prompt (optional), reasoning_mode no_think (default, direct answer) or think (returns a separate reasoning trace), max_tokens (default 1024), temperature (default 1.0, NVIDIA's recommended value), top_p (default 0.95), safety checker on by default. Returns output text, finish_reason, token usage and, in think mode, the reasoning trace. Billed per 1,000 tokens (input plus output), minimum one unit per request.

**Tags**: vision, multimodal, reasoning



## Pricing

Your request will cost **$0.01** per 1,000 tokens (input plus output), with a minimum of **$0.01** per request.

For more details, see [fal.ai pricing](https://fal.ai/pricing).

## API Information

This model can be used via our HTTP API or more conveniently via our client libraries.
See the input and output schema below, as well as the usage examples.


### Input Schema

The API accepts the following input parameters:


- **`prompt`** (`string`, _required_):
  Text prompt to send to the model.
  - Examples: "Summarize the key ideas of hybrid Mamba-Transformer models."

- **`system_prompt`** (`string`, _optional_):
  Optional system prompt to steer the model.
  - Examples: "You are a concise assistant."

- **`reasoning_mode`** (`ReasoningModeEnum`, _optional_):
  `think` emits an explicit reasoning trace before the final answer; `no_think` returns a direct answer. Default value: `"no_think"`
  - Default: `"no_think"`
  - Options: `"think"`, `"no_think"`

- **`max_tokens`** (`integer`, _optional_):
  Maximum number of tokens to generate. Default value: `1024`
  - Default: `1024`
  - Range: `1` to `32000`

- **`temperature`** (`float`, _optional_):
  Sampling temperature. NVIDIA recommends 1.0 for all tasks. Default value: `1`
  - Default: `1`
  - Range: `0` to `2`

- **`top_p`** (`float`, _optional_):
  Nucleus sampling probability mass. Default value: `0.95`
  - Default: `0.95`

- **`enable_safety_checker`** (`boolean`, _optional_):
  Whether to enable the safety checker. Default value: `true`
  - Default: `true`

- **`image_url`** (`string`, _required_):
  URL of the image to reason about. RGB JPEG or PNG.
  - Examples: "https://storage.googleapis.com/falserverless/example_inputs/dog.png"



**Required Parameters Example**:

```json
{
  "prompt": "Summarize the key ideas of hybrid Mamba-Transformer models.",
  "image_url": "https://storage.googleapis.com/falserverless/example_inputs/dog.png"
}
```

**Full Example**:

```json
{
  "prompt": "Summarize the key ideas of hybrid Mamba-Transformer models.",
  "system_prompt": "You are a concise assistant.",
  "reasoning_mode": "no_think",
  "max_tokens": 1024,
  "temperature": 1,
  "top_p": 0.95,
  "enable_safety_checker": true,
  "image_url": "https://storage.googleapis.com/falserverless/example_inputs/dog.png"
}
```


### Output Schema

The API returns the following output format:

- **`output`** (`string`, _required_):
  Generated text response.
  - Examples: "Hello! How can I help?"

- **`reasoning`** (`string`, _optional_):
  Reasoning trace, returned only in `think` mode.

- **`finish_reason`** (`string`, _optional_):
  Reason generation stopped. Default value: `"stop"`
  - Default: `"stop"`
  - Examples: "stop", "length"

- **`usage`** (`UsageInfo`, _required_):
  Token usage for the request.
  - Examples: {"input_tokens":24,"output_tokens":87}



**Example Response**:

```json
{
  "output": "Hello! How can I help?",
  "finish_reason": "stop",
  "usage": {
    "input_tokens": 24,
    "output_tokens": 87
  }
}
```


## Usage Examples

### cURL

```bash
curl --request POST \
  --url https://fal.run/nvidia/nemotron-3-5-super/vision \
  --header "Authorization: Key $FAL_KEY" \
  --header "Content-Type: application/json" \
  --data '{
     "prompt": "Summarize the key ideas of hybrid Mamba-Transformer models.",
     "image_url": "https://storage.googleapis.com/falserverless/example_inputs/dog.png"
   }'
```

### Python

Ensure you have the Python client installed:

```bash
pip install fal-client
```

Then use the API client to make requests:

```python
import fal_client

def on_queue_update(update):
    if isinstance(update, fal_client.InProgress):
        for log in update.logs:
           print(log["message"])

result = fal_client.subscribe(
    "nvidia/nemotron-3-5-super/vision",
    arguments={
        "prompt": "Summarize the key ideas of hybrid Mamba-Transformer models.",
        "image_url": "https://storage.googleapis.com/falserverless/example_inputs/dog.png"
    },
    with_logs=True,
    on_queue_update=on_queue_update,
)
print(result)
```

### JavaScript

Ensure you have the JavaScript client installed:

```bash
npm install --save @fal-ai/client
```

Then use the API client to make requests:

```javascript
import { fal } from "@fal-ai/client";

const result = await fal.subscribe("nvidia/nemotron-3-5-super/vision", {
  input: {
    prompt: "Summarize the key ideas of hybrid Mamba-Transformer models.",
    image_url: "https://storage.googleapis.com/falserverless/example_inputs/dog.png"
  },
  logs: true,
  onQueueUpdate: (update) => {
    if (update.status === "IN_PROGRESS") {
      update.logs.map((log) => log.message).forEach(console.log);
    }
  },
});
console.log(result.data);
console.log(result.requestId);
```


## Additional Resources

### Documentation

- [Model Playground](https://fal.ai/models/nvidia/nemotron-3-5-super/vision)
- [API Documentation](https://fal.ai/models/nvidia/nemotron-3-5-super/vision/api)
- [OpenAPI Schema](https://fal.ai/api/openapi/queue/openapi.json?endpoint_id=nvidia/nemotron-3-5-super/vision)

### fal.ai Platform

- [Platform Documentation](https://fal.ai/docs/documentation)
- [Python Client](https://fal.ai/docs/api-reference/client-libraries/python)
- [JavaScript Client](https://fal.ai/docs/api-reference/client-libraries/javascript)

### Other agent-readable surfaces

This file covers one model. To find anything else:

- [Platform overview](https://fal.ai/llms.txt): Entry points and representative endpoint IDs
- [Documentation index](https://fal.ai/docs/llms.txt): Every documentation page
- [Full documentation text](https://fal.ai/docs/llms-full.txt): The whole documentation inlined
- Any other model: `https://fal.ai/models/<endpoint-id>/llms.txt`