# Nova Micro

Nova Micro is optimized for speed. Amazon limited this model to text processing so it can prioritize throughput and low per-token cost relative to Nova models that include vision.

- **Model ID:** `amazon/nova-micro`
- **Type:** chat
- **Providers:** bedrock
- **Context window:** 128,000
- **Maximum output tokens:** 8,192
- **Pricing:** $0.035/1M input tokens, $0.14/1M output tokens
- **Canonical page:** https://vercel.com/ai-gateway/models/nova-micro

## Supported parameters

Detailed capability metadata has not been reported for this model.

## Example

```ts
import { streamText } from 'ai'

const result = streamText({
  model: 'amazon/nova-micro',
  prompt: 'Why is the sky blue?'
})
```

## About

Nova Micro launched alongside the rest of the first-generation Nova family, but its design philosophy differs. Nova Lite and Nova Pro layer in image and video understanding. Nova Micro drops multimodal support entirely. The result is a model that does one thing, text processing, at high speed and low cost.

The tradeoff is intentional. Removing vision processing frees up architecture for generation throughput within a context window of 128K tokens. Even high-volume classification or tagging pipelines have low per-request costs.

Many teams deploy Nova Micro as the default tier in a routing architecture. Straightforward text requests (classification, entity extraction, simple Q&A, short summaries) go to Micro. Only when a request involves images, requires deep reasoning, or exceeds the context of 128K tokens does it escalate to Lite, Pro, or a second-generation model. This pattern keeps average cost per request low while covering the full range of task complexity.

## What to consider

If streaming throughput matters, test Nova Micro early. It's built for high text throughput within AI Gateway.

## When to use

### Best For

- Chatbots and conversational interfaces where response speed affects UX
- Text classification, sentiment analysis, and entity extraction at scale
- Autocomplete and inline suggestion features
- The default tier in a model-routing architecture, handling routine text requests at minimal cost

### Consider Alternatives When

- Inputs include images, documents, or video. Switch to Nova Lite or Nova Pro
- The task demands structured multi-step reasoning or agentic tool use. Nova 2 Lite is better equipped
- You need more than 128K tokens of context for long documents or extended conversations

## Best for

- **Chatbots and conversational interfaces:** Interactive experiences where response speed affects UX
- **Text classification at scale:** Sentiment analysis and entity extraction across high request volumes
- **Autocomplete and suggestions:** Inline features where generation speed matters
- **Default routing tier:** Handle routine text requests at minimal cost in a model-routing architecture

## Consider alternatives

- **Multimodal inputs required:** Switch to Nova Lite or Nova Pro when inputs include images, documents, or video
- **Structured multi-step reasoning:** Nova 2 Lite is better equipped for agentic tool use and deeper reasoning
- **Context beyond 128K tokens:** Long documents or extended conversations require a model with a larger context window

## Frequently asked questions

### Why would I choose Nova Micro over Nova Lite for text tasks?

Nova Micro is priced below Nova Lite's multimodal rate and is tuned for speed on pure text. If you never send images or video, Micro is usually the cheaper fit.

### Can Nova Micro handle long documents?

Keep the prompt within 128K tokens. If you exceed that, split the document, summarize in chunks, or switch to Nova 2 Lite for a 1M-token window.

### Is Nova Micro good for structured output like JSON?

Yes. It follows instructions well for classification, tagging, and structured extraction. Its speed makes it especially efficient for pipelines that process many short requests.

### What is the maximum output length?

Nova Micro generates up to 8.2K tokens per response.

### How does Nova Micro perform on reasoning benchmarks?

Nova Micro isn't designed for complex reasoning. It excels at speed and cost efficiency for routine language tasks. For reasoning-heavy workloads, consider Nova 2 Lite or Nova Pro.

### Do I need separate AWS credentials?

No. AI Gateway handles authentication with Amazon Bedrock. You only need a gateway API key or OIDC token.

## Links

- [Model page](https://vercel.com/ai-gateway/models/nova-micro)
- [AI Gateway documentation](https://vercel.com/docs/ai-gateway)
- [Provider model documentation](https://docs.aws.amazon.com/nova/latest/userguide/what-is-nova.html)
- [Provider pricing](https://aws.amazon.com/bedrock/pricing/)