Models·Amazon

Nova Pro 1.0: Amazon Model Pricing and Limits

amazon/nova-pro-v1

Amazon’s multimodal model for text and image tasks, with tool support and a 300,000-token context window.

$0.80Input, per million tokens
$3.20Output, per million tokens
300KContext window
5KLongest answer

300K tokens of context — about 225,000 words, on a scale that ends at 2M.

What it can do

  • Vision
  • Tool use
  • Takes text
  • Takes image
  • Gives text

Released 5 Dec 2024 · checked 3 Oct 2026.

About it

Nova Pro 1.0 is an Amazon model that accepts text and images, returns text, and can use tools. It can read up to 300,000 tokens in one request and produce up to 5,120 tokens in one answer. Input costs $0.8 per million tokens and output costs $3.2 per million tokens.

Review

In a line

Nova Pro 1.0 is a multimodal model from Amazon. It works with text and images, then produces text responses.

What it's good at

This model suits tasks that combine written material and pictures. Its tool support allows software using it to call tools while handling a request.

For a broader list of Amazon offerings, see Amazon models. Its image input also places it among vision language models.

Price in plain terms

Input costs $0.8 per million tokens. Output costs $3.2 per million tokens. Tokens are small pieces of text counted for billing.

At the stated usage levels, reading a 500-word page costs about 0.052 cents. One chat turn with about 1,000 input tokens and 500 output tokens costs about 0.24 cents. One thousand such turns cost about $2.40.

Context and limits

Nova Pro 1.0 can take up to 300,000 tokens in one context, so a single request can include a large amount of text and image input.

Each answer can contain at most 5,120 tokens. The model's output is text, even when an image is included as input.

Ways to use it

Use it in software that supports Amazon's model for requests containing text, images, or both. Tool support lets that software give the model access to tools during a task.

Its 300,000-token context can hold substantial material.

Verdict

Choose Nova Pro 1.0 when you need one model to handle written input and images, with optional tool use and a large context. The listed token prices make repeated, moderate chats relatively inexpensive.

Look elsewhere if you need answers longer than 5,120 tokens at once, or if your work requires an output format other than text.

Verdict

Nova Pro 1.0 is a sensible fit for teams building text-and-image features around Amazon’s model family. Its large context can hold substantial material, while the per-token rates keep the sample chat workload near $2.40 for 1,000 turns. The main trade-off is output length: a single response stops at 5,120 tokens. It also returns text only, so it is not suited to workflows that need generated images or other output types.

  • Accepts both text and image input
  • Supports tool use
  • 300,000-token context window
  • Costs about $2.40 for 1,000 sample chat turns
  • Maximum response is 5,120 tokens
  • Responses are text only
  • Output tokens cost more than input tokens
  • Pick it ifText-and-image tasks; Software that can use model tools; Large-context requests
  • Skip it ifYou need responses longer than 5,120 tokens in one answer; You need non-text output

Pricing

Billing is per million tokens: $0.8 for input and $3.2 for output. A 500-word page costs about 0.052 cents to read; 1,000 sample chat turns cost about $2.40.

Questions

Who makes Nova Pro 1.0?

Amazon makes Nova Pro 1.0.

What can Nova Pro 1.0 read?

It accepts text and images as input.

What does it return?

It returns text.

How much context can it handle?

Its context window is 300,000 tokens.

What is the maximum answer length?

A single answer can contain up to 5,120 tokens.

Does it support tools?

Yes. The model is listed as supporting tools.

What does it cost?

Input is $0.8 per million tokens and output is $3.2 per million tokens.

On these boards

More from Amazon