We use cookies. Cookies and browser storage keep you signed in and remember your preferences. With your permission we also measure how the site is used. You can any time. See our cookie policy.

New here? See how it works A 1-minute tour of the playground, models and your API key.

DeepSeek V4 Pro 0423 vs Nemotron 3.5 Super VL: API price, context & features · api14

DeepSeek: DeepSeek V4 Pro 0423 vs NVIDIA: Nemotron 3.5 Super VL

A side-by-side comparison of DeepSeek: DeepSeek V4 Pro 0423 ($5.60 / 1M tokens) and NVIDIA: Nemotron 3.5 Super VL ($0.52 / 1M tokens) using current USD prices and capabilities from the api14 catalog. Both run through the same OpenAI-compatible API.

Specifications

	DeepSeek: DeepSeek V4 Pro 0423	NVIDIA: Nemotron 3.5 Super VL
Provider	DeepSeek	NVIDIA
Input	$5.60 / 1M tokens	$0.52 / 1M tokens
Output	$11.20 / 1M tokens	$1.75 / 1M tokens
Cache read	$0.47 / 1M tokens	$0.07 / 1M tokens
Context window	1,048,576	1,048,576
Max output	393,216	943,718
Input	text	text, image, video
Output	text	text
Tool calling
Structured output
Reasoning control
Added	Apr 24, 2026	Oct 15, 2026

Key differences

NVIDIA: Nemotron 3.5 Super VL is 91% cheaper on input tokens.
NVIDIA: Nemotron 3.5 Super VL is 84% cheaper on output tokens.
A request with 1,000 input and 500 output tokens costs $0.01 with DeepSeek: DeepSeek V4 Pro 0423 and $0.00140 with NVIDIA: Nemotron 3.5 Super VL.

In depth

DeepSeek: DeepSeek V4 Pro 0423 is a text model from DeepSeek. NVIDIA: Nemotron 3.5 Super VL is a text model from NVIDIA. Both sit in the text category on api14, and both were released by their providers on the dates listed for them, but they are built for different kinds of input.

The two models share the same context length of 1048576 tokens, so neither gives you a longer working window than the other. They also expose nearly the same set of sampling and control parameters, including temperature, top_p, top_k, min_p, frequency and presence penalties, repetition penalty, seed, stop, logit_bias, logprobs and top_logprobs, response_format, structured_outputs, tools and tool_choice, plus the reasoning related options. The real split is in input modalities. DeepSeek V4 Pro 0423 accepts text only and returns text. Nemotron 3.5 Super VL accepts text, image and video as input and returns text, which makes it the one to reach for when your prompt includes a screenshot, a diagram, a scanned page or a video clip. Output on both sides is text.

For day to day work that is pure language, either one is a reasonable default, and the tie on context length means you can feed very long documents to both. When your task is summarising a long report, drafting or refactoring code, extracting structured fields from text, or running a tool calling loop, DeepSeek V4 Pro 0423 covers it, and it carries the full set of parameters listed for it. When your task involves reading something visual, such as describing what happens in a video, pulling text out of an image, or answering questions about a chart, Nemotron 3.5 Super VL is the one that can see the input at all, because DeepSeek V4 Pro 0423 takes text only. Note that Nemotron lists max_tokens rather than max_completion_tokens, so if your client sends max_completion_tokens you should switch that field to max_tokens for that model.

Calling either model through api14 works the same way, because the API is OpenAI compatible. You point your existing OpenAI style client at the api14 base URL, send your usual chat completions request, and change only the model id. Use "deepseek/deepseek-v4-pro" for the DeepSeek model and "nvidia/nemotron-3.5-super-vl" for the NVIDIA model. Everything else in the request body, your messages, your sampling parameters, your tools and your response format, stays as it is, aside from the max_tokens note above.

On price, the NVIDIA model is the cheaper of the two for both input and output, so if your workload is text only and cost matters, Nemotron 3.5 Super VL is the more economical call, and it also happens to cover the image and video cases. The case for DeepSeek V4 Pro 0423 is the text only workload where you want the DeepSeek model specifically, and it remains the natural pick whenever the input is plain text and you prefer that provider. Check the live price table on this page for the current numbers.

Switch with one string

Both models use the same endpoint and request shape. Change the model value from "deepseek/deepseek-v4-pro" to "nvidia/nemotron-3.5-super-vl" and prompts, streaming and tools keep working.

DeepSeek: DeepSeek V4 Pro 0423 details & playground →NVIDIA: Nemotron 3.5 Super VL details & playground →

import OpenAI from "openai";

const client = new OpenAI({
  baseURL: "https://api14.com/api/v1",
  apiKey: process.env.API14_API_KEY,
});

const result = await client.chat.completions.create({
  "model": "nvidia/nemotron-3.5-super-vl",
  "messages": [
    {
      "role": "user",
      "content": "Explain vector databases in two sentences."
    }
  ]
});
console.log(result.choices[0].message.content);

DeepSeek: DeepSeek V4 Pro 0423 vs NVIDIA: Nemotron 3.5 Super VL

Specifications

Key differences

In depth

Switch with one string

More comparisons