impulse

Nova Micro

Amazon Nova Micro 1.0 is a text-only model that delivers the lowest latency responses in the Amazon Nova family of models at a very low cost.

API id nova-micro-v1Released 5 Dec 2024 (Dec 2024)
Input / 1M
$0.035
at Amazon Nova
Output / 1M
$0.14
at Amazon Nova
Context window
128K
tokens
Max output
10K
tokens per reply
Hosts
3
29 listings
Released
Dec 2024
5 Dec 2024

Pricing

List prices at Amazon Nova, in US dollars per million tokens.

TierInput / 1MOutput / 1M
Standard$0.035$0.14

What it costs a month

At Amazon Nova list prices, 30 days.

  • Chat assistant5,000 requests a day, 2K in / 0.5K out$21.00
  • RAG over documents2,000 requests a day, 8K in / 0.6K out$21.84
  • Coding agent1,000 requests a day, 40K in / 2K out$50.40
  • Overnight batch20,000 requests a day, 4K in / 0.4K out$118
Use your own numbers

Where to run it

3 hosts serve Nova Micro. The comparison is on blended price, 3 parts input to 1 part output.

HostIn / out per 1M
nova-micro-v1
$0.035 / $0.14
amazon.nova-micro-v1:0
$0.035 / $0.14
nova-micro
$0.035 / $0.14
Show 26 more listings: regions, snapshots and variants
Amazon Bedrock
me-central-1
amazon.nova-micro-v1:0
$0.035 / $0.14
us-east-1
amazon.nova-micro-v1:0
$0.035 / $0.14
us-east-2
amazon.nova-micro-v1:0
$0.035 / $0.14
us-west-2
amazon.nova-micro-v1:0
$0.035 / $0.14
us
us.amazon.nova-micro-v1:0
$0.035 / $0.14
ap-southeast-2
amazon.nova-micro-v1:0
$0.037 / $0.148
+6% vs headline
apac
apac.amazon.nova-micro-v1:0
$0.037 / $0.148
+6% vs headline
eu-north-1
amazon.nova-micro-v1:0
$0.038 / $0.152
+9% vs headline
eu
eu.amazon.nova-micro-v1:0
$0.038 / $0.152
+9% vs headline
eu-south-2
amazon.nova-micro-v1:0
$0.039 / $0.156
+11% vs headline
eu
eu.amazon.nova-micro-v1:0
$0.039 / $0.156
+11% vs headline
eu-west-1
amazon.nova-micro-v1:0
$0.04 / $0.16
+14% vs headline
eu
eu.amazon.nova-micro-v1:0
$0.04 / $0.16
+14% vs headline
ap-northeast-2
amazon.nova-micro-v1:0
$0.041 / $0.164
+17% vs headline
ap-south-1
amazon.nova-micro-v1:0
$0.041 / $0.164
+17% vs headline
us-gov-west-1
amazon.nova-micro-v1:0
$0.042 / $0.168
+20% vs headline
ap-northeast-1
amazon.nova-micro-v1:0
$0.042 / $0.168
+20% vs headline
il-central-1
amazon.nova-micro-v1:0
$0.044 / $0.176
+26% vs headline
eu-central-1
amazon.nova-micro-v1:0
$0.046 / $0.184
+31% vs headline
eu
eu.amazon.nova-micro-v1:0
$0.046 / $0.184
+31% vs headline
ap-southeast-1
amazon.nova-micro-v1:0
$0.047 / $0.188
+34% vs headline
eu-west-2
amazon.nova-micro-v1:0
$0.049 / $0.196
+40% vs headline
eu-west-3
amazon.nova-micro-v1:0
$0.052 / $0.208
+49% vs headline
eu
eu.amazon.nova-micro-v1:0
$0.052 / $0.208
+49% vs headline
eu-south-1
amazon.nova-micro-v1:0
$0.056 / $0.224
+60% vs headline
eu
eu.amazon.nova-micro-v1:0
$0.056 / $0.224
+60% vs headline

Capabilities

Core

  • Streaming
  • Function calling
  • JSON mode
  • Structured outputs

Multimodal

  • Vision
  • Image input
  • Audio
  • Image generation
  • Video

Advanced

  • Reasoning
  • Tool calling
  • MCP
  • Prompt caching
  • Batch API
  • Embeddings
Takes in
Text
Produces
Text
Features
Prompt caching

Cheaper alternatives

Current models that cost less and keep most of what this one can do: tools, vision, reasoning, structured output, caching and enough context.

Same model, cheaper host

No host sells it for less than Amazon Nova.

Smaller Amazon models

No cheaper Amazon model keeps its key capabilities.

Other frontier labs

No cheaper match from another frontier lab.

Open-weight picks

  • Qwen3.7 Flash Variant1
    Alibaba Qwen · on EmpirioLabs AI
    $0.028 / $0.11
    Has every key capability
    21% less
  • Llama 3.2 3B
    Meta · on LlamaGate
    $0.04 / $0.08
    Lacks caching
    18% less
  • Gemma3 4B
    Google · on LlamaGate
    $0.03 / $0.08
    Lacks caching
    31% less
  • Ministral 3B
    Mistral AI · on Azure AI Foundry
    $0.04 / $0.04
    Lacks caching
    35% less

Call it

Using Amazon Nova's API.

Python
from litellm import completion

response = completion(
    model="amazon_nova/nova-micro-v1",
    messages=[{"role": "user", "content": "Hello"}],
)
print(response.choices[0].message.content)

Quick answers

How much does Nova Micro cost?

On Amazon Nova, Nova Micro costs $0.035 per million input tokens and $0.14 per million output tokens.

Which APIs offer Nova Micro?

Nova Micro runs on 3 APIs: Amazon Nova, Amazon Bedrock and Vercel AI Gateway.

What is Nova Micro's context window?

128K tokens, with up to 10K tokens in one reply.

Is Nova Micro being retired?

No retirement date is scheduled for it on any API we track.

What are cheaper alternatives to Nova Micro?

Models that cost less and keep most of what it can do: Qwen3.7 Flash Variant1 on EmpirioLabs AI (21% less). Savings are on a mix of 3 input tokens to 1 output token.

When was Nova Micro released?

Amazon released it on 5 Dec 2024.

What is Nova Micro's API model id?

Use nova-micro-v1 with Amazon Nova. Other hosts use their own ids: amazon.nova-micro-v1:0 on Amazon Bedrock, nova-micro on Vercel AI Gateway.

Sources

Raw catalog entry for nova-micro-v1 on Amazon Nova
{
  "input_cost_per_token": 3.5e-8,
  "litellm_provider": "amazon_nova",
  "max_input_tokens": 128000,
  "max_output_tokens": 10000,
  "mode": "chat",
  "output_cost_per_token": 1.4e-7,
  "supports_function_calling": true,
  "supports_prompt_caching": true,
  "supports_response_schema": true
}