Amazon models on Amazon Bedrock
14 Amazon models available through Bedrock, with price per 1M tokens, context window and region coverage.
| Model | Context | Input | Output | Regions |
|---|---|---|---|---|
| Amazon Nova Multimodal Embeddings amazon.nova-2-multimodal-embeddings-v1:0 | — | — | — | On-demand 1 Inference profile only 0 |
| Nova 2 Lite amazon.nova-2-lite-v1:0 | — | $0.300 | $2.50 | On-demand 0 Inference profile only 25 |
| Nova 2 Sonic amazon.nova-2-sonic-v1:0 | — | — | — | On-demand 4 Inference profile only 0 |
| Nova Canvas amazon.nova-canvas-v1:0 | — | — | — | On-demand 3 Inference profile only 0 |
| Nova Lite amazon.nova-lite-v1:0 | 300K | $0.060 | $0.240 | On-demand 8 Inference profile only 16 |
| Nova Micro amazon.nova-micro-v1:0 | 128K | $0.035 | $0.140 | On-demand 3 Inference profile only 17 |
| Nova Pro amazon.nova-pro-v1:0 | 300K | $0.800 | $3.20 | On-demand 4 Inference profile only 18 |
| Nova Reel amazon.nova-reel-v1:0 | — | — | — | On-demand 3 Inference profile only 0 |
| Nova Reel amazon.nova-reel-v1:1 | — | — | — | On-demand 1 Inference profile only 0 |
| Rerank 1.0 amazon.rerank-v1:0 | — | — | — | On-demand 4 Inference profile only 0 |
| Titan Embeddings G1 - Text amazon.titan-embed-text-v1 | 8K | $0.100 | — | On-demand 4 Inference profile only 0 |
| Titan Multimodal Embeddings G1 amazon.titan-embed-image-v1 | — | — | — | On-demand 10 Inference profile only 0 |
| Titan Text Embeddings v2 amazon.titan-embed-g1-text-02 | — | — | — | On-demand 2 Inference profile only 0 |
| Titan Text Embeddings V2 amazon.titan-embed-text-v2:0 | 8K | — | — | On-demand 19 Inference profile only 0 |
Common questions
- Which Amazon model is cheapest on Amazon Bedrock?
- Nova Micro at $0.035 per 1M input tokens in us-east-1, against $0.140 per 1M output tokens.
- Which Amazon model is available in the most AWS regions?
- Nova 2 Lite, callable in 25 of the 31 regions tracked here — 0 on-demand and 25 through a cross-region inference profile.
- Which Amazon model has the largest context window on Bedrock?
- Nova Lite, with a 300K-token context window.