Skip to main content
Watch: I Built an AI App Without Leaving Databricks

Amazon - Closed weights

Amazon Nova 2 Lite pricing and specs

Amazon Nova 2 Lite is a closed-weights model from Amazon that costs $0.30 per 1M input tokens and $2.50 per 1M output tokens on the first-party API, with a 1M-token context window.

Last verified from aws.amazon.com. All sources

Input / 1M
$0.30
Output / 1M
$2.50
Context
1M tokens
Weights
Closed

Pricing per 1M tokens

ItemPrice / 1M tokens
Input$0.30
Output$2.50
Cache read$0.075
Batch inputBatch requests$0.15
Batch outputBatch requests$1.25

Note: Prices are US East (N. Virginia), Standard tier, Global cross-Region inference (global.amazon.nova-2-lite-v1:0). Geo cross-Region (us. profile) Standard is $0.33 input / $2.75 output and Geo batch is $0.165 / $1.375. us-east-1 has no in-Region option for this model. Other tiers (Global): Priority $0.525 / $4.375, Flex $0.15 / $1.25. The $0.075 cache-read price is printed in the page's model-customization On Demand Inference table, which states custom models are priced the same as base models; the page footnote also says cache reads are 75% less than the on-demand input price. No cache-write price is listed for Nova. Prompt caching tops out at 20K tokens on Nova models. The model card says max output 64K; the Nova 2 developer guide says 65,536 tokens. The pricing page shows US East (Ohio) by default. These N. Virginia figures come from the page's own regional price data for US East (N. Virginia). Web grounding is billed separately at $30 per 1K requests.

Specs

Context window
1,000,000 tokens
Max output
65,536 tokens
Release date
Provider
Amazon
API model id
amazon.nova-2-lite-v1:0
Vercel AI Gateway id
amazon/nova-2-lite
Weights
Closed weights

What Amazon Nova 2 Lite costs in practice

Worked from the standard list rates above, with the math shown so you can swap in your own token counts.

One agent session

1M input tokens and 100K output tokens, no caching.

  • Input1M x $0.30 = $0.30
  • Output0.1M x $2.50 = $0.25

$0.55per session

A chat app

10,000 requests a day at 2K input and 500 output tokens each.

  • Input20M x $0.30 = $6.00
  • Output5M x $2.50 = $12.50

$18.50per day

$555.00 per 30-day month

The same session at a 90% cache hit

90% of the input is read from the prompt cache. The one-time cache write premium is not included.

  • Cached input0.9M x $0.075 = $0.0675
  • Uncached input0.1M x $0.30 = $0.03
  • Output0.1M x $2.50 = $0.25

$0.3475per session

Compare with

ModelInput / 1MOutput / 1M
Amazon Nova 2 Lite (this page)$0.30$2.50
Dola Seed 2.1 TurboByteDance Seed$0.50$2.50
Gemini 2.5 FlashGoogle$0.30$2.50
Gemini 3.5 Flash-LiteGoogle$0.30$2.50
Grok 4.20xAI$1.25$2.50
Amazon Nova ProAmazon$0.80$3.20
Amazon Nova LiteAmazon$0.06$0.24
Amazon Nova MicroAmazon$0.035$0.14
Amazon Nova PremierAmazon$2.50$12.50

Related reading

FAQ

How much does Amazon Nova 2 Lite cost per million tokens?

Amazon Nova 2 Lite costs $0.30 per 1M input tokens and $2.50 per 1M output tokens on Amazon's first-party API, per aws.amazon.com on October 17, 2026. Per 1M tokens, cache reads cost $0.075, and the batch rate is $0.15 input and $1.25 output.

What is Amazon Nova 2 Lite's context window?

Amazon Nova 2 Lite has a 1,000,000-token context window and returns up to 65,536 output tokens per response.

Are Amazon Nova 2 Lite's weights open?

No. Amazon Nova 2 Lite is a closed-weights model. You call it through Amazon's API as amazon.nova-2-lite-v1:0 or through Vercel AI Gateway as amazon/nova-2-lite.

When was Amazon Nova 2 Lite released?

Amazon released Amazon Nova 2 Lite on December 2, 2025.

Sources

Last verified from aws.amazon.com.

Newsletter

Get model price changes as they ship

One email per week. New model releases, price cuts and what they mean for your bill.