AI Model Watch
Models / Amazon

Amazon Nova Multimodal Embeddings

GA

Unified multimodal embedding model (text, documents, images, video, audio in; vector out — no output-token price). Context up to 8K tokens; video/audio segments up to 30s with built-in chunking; sync + async (StartAsyncInvoke) APIs. Output dimensions 3,072 / 1,024 / 384 / 256 (docs examples default 3,072). The $0.135/M figure is the TEXT input rate (us-east-1 on-demand, AWS Price List API: $0.000135/1K); other inputs billed separately — standard image $0.00006/image, document image $0.0006/image, audio $0.00014/s, video $0.0007/s (batch = 50% of each). Model card lifecycle: Active, EOL N/A.

Amazon Nova Multimodal Embeddings by Amazon is tracked live in our LLM catalog, with a 8K (8192-token) context window. It is generally available (GA).

Last verified: 13 Aug 2026 · sourced from official provider documentation

Provider
Amazon
Status
GA
Input price
$0.135 / 1M tokens
Output price
—
Cached input
—
Blended price
—
Context window
8192 tokens (8K)
Max output
—
Embedding dimensions
3072-dim output vectors
Modality
text, image, video, audio
Open weights
No — API only
Knowledge cutoff
—
Released
28 Oct 2025
API string
amazon.nova-2-multimodal-embeddings-v1:0

Source: Amazon official documentation ↗

See also

Explore more

All Amazon models→

Prices, context windows and lifecycle status for every Amazon model we track.

Compare models head-to-head→

Side-by-side price, context and status for the models developers weigh against each other.

FAQ

Amazon Nova Multimodal Embeddings — questions & answers

What is the context window of Amazon Nova Multimodal Embeddings?

Amazon Nova Multimodal Embeddings has a 8192-token context window (8K).

Is Amazon Nova Multimodal Embeddings deprecated?

No — Amazon Nova Multimodal Embeddings is generally available (GA) and not currently scheduled for deprecation or retirement in our tracker.

Does Amazon Nova Multimodal Embeddings support image or vision input?

Yes — Amazon Nova Multimodal Embeddings accepts image input. Listed modalities: text, image, video, audio.

Can I self-host Amazon Nova Multimodal Embeddings?

No — Amazon Nova Multimodal Embeddings is proprietary. Its weights are not publicly released, so it is available only through Amazon's API (or hosted partners), not for self-hosting.

What is the API model string for Amazon Nova Multimodal Embeddings?

The API model identifier for Amazon Nova Multimodal Embeddings is "amazon.nova-2-multimodal-embeddings-v1:0" when calling Amazon's API.