AWS Bedrock via VerticalAPI

Supported models

AWS Bedrock models routed by VerticalAPI

Pass the model ID below as model in any OpenAI-compatible request. New AWS Bedrock models are typically supported within 24h of release.

Model ID	Name	Context	Pricing (provider)
`anthropic.claude-sonnet-4-5-v1:0`	Claude Sonnet 4.5 (Bedrock)	200K	Anthropic Bedrock pricing
`meta.llama3-3-70b-instruct-v1:0`	Llama 3.3 70B (Bedrock)	128K	AWS Llama pricing
`mistral.mistral-large-2407-v1:0`	Mistral Large 2 (Bedrock)	128K	AWS Mistral pricing
`amazon.titan-text-premier-v1:0`	Titan Text Premier	32K	AWS Titan pricing

Pricing reflects AWS Bedrock's rates — you pay AWS Bedrock directly. VerticalAPI adds zero markup on tokens.

Quickstart

5-line AWS Bedrock call via VerticalAPI

Drop-in replacement for the OpenAI SDK. Works with the OpenAI Python client, Node, Go, curl — anything that speaks HTTP.

                aws-bedrock_quickstart.py
                Python
            
from openai import OpenAI

client = OpenAI(
    base_url="https://api.verticalapi.com/v1",
    api_key="vapi_...",
    default_headers={"X-Provider-Key": "AWS access key..."}
)

response = client.chat.completions.create(
    model="anthropic.claude-sonnet-4-5-v1:0",  # AWS Bedrock
    messages=[{"role": "user", "content": "Hello"}]
)
print(response.choices[0].message.content)

AWS Bedrock and VerticalAPI

What VerticalAPI does and does not do with AWS Bedrock

Not routed today

VerticalAPI does not proxy AWS Bedrock. This page documents AWS Bedrock's own models and pricing so you can compare it; call AWS Bedrock directly to use it.

What it does route

11 providers behind one OpenAI-compatible endpoint: Anthropic, Cerebras, DeepInfra, Fireworks, Google AI Studio, Groq, Mistral, OpenAI, OpenRouter, Together, xAI. Switchable per request, same SDK.

No token markup on those

Bring your own key and pay the provider directly. VerticalAPI's revenue is the gateway subscription, not a share of your tokens.

Prompts are never stored

Each request is recorded as metadata only — vertical, model, token counts, latency, status. There is no column for prompt or completion text.

Best for

Where AWS Bedrock shines

AWS-native data residency VPC-only inference BAA / HIPAA pipelines consolidated AWS billing

Frequently asked questions

What is AWS Bedrock and what models do they offer?

AWS Bedrock is Amazon's fully managed foundation model service. The 2026 catalog includes Anthropic Claude 4.5 (Sonnet/Opus/Haiku), Meta Llama 3.3, Mistral Large 2, Cohere Command R+, AI21 Jamba 1.5, Stability SD3, plus Amazon's Nova (Pro, Lite, Micro, Canvas, Reel) and Titan families. All models share one Converse API, IAM access control, CloudWatch logging, VPC endpoints and regional data residency.

How much does AWS Bedrock cost in 2026?

Bedrock charges each vendor's list price: Claude Sonnet 4.5 at $3/$15, Llama 3.3 70B at $0.72/$0.72, Mistral Large 2 at $3/$9, Cohere Command R+ at $2.50/$10, Nova Pro at $0.80/$3.20. Provisioned throughput offers discounted rates for committed capacity. Plus AWS data transfer and CloudWatch fees. You pay AWS Bedrock directly, on AWS Bedrock's own account.

Can I use AWS Bedrock through VerticalAPI?

Create an IAM user with bedrock:InvokeModel permissions, generate access keys, paste them into VerticalAPI (with region), then point the OpenAI SDK at https://api.verticalapi.com/v1. VerticalAPI signs SigV4 requests for you and translates OpenAI chat completions into Bedrock Converse calls. Billing remains on your AWS invoice with full Cost Explorer attribution.

What is AWS Bedrock best for compared to alternatives?

Bedrock wins for AWS-native enterprises: single IAM-governed gateway, VPC isolation, customer-managed KMS keys, CloudTrail audit, and regional data residency (EU, US, APAC, GovCloud). Compared to direct vendor APIs it adds AWS lock-in but consolidates billing and security. Compared to Vertex AI it has broader third-party model selection. Not the cheapest path to a single model — direct vendor or open-weight inference is usually less expensive.

Where is AWS Bedrock hosted / data privacy?

Bedrock is available in 15+ AWS regions including us-east-1, us-west-2, eu-west-1, eu-central-1, ap-northeast-1 and GovCloud. Inputs and outputs stay in the chosen region and are not used to train models. CMK encryption, PrivateLink, VPC endpoints and CloudTrail provide enterprise-grade governance. Your contract and data terms stay between you and AWS Bedrock.

Limitations and trade-offs

Pricing is at vendor list — never cheaper than going direct, sometimes slightly more due to AWS surcharges.
Model availability and versions lag behind direct vendor APIs by weeks or months.
Latency is region-bound — you must pick a region close to your users and to the model's availability.
Bedrock Converse API has subtle differences from direct vendor APIs (e.g. system prompts, tool use formatting).
Provisioned throughput requires committed monthly spend and is wasteful for spiky workloads.

Where AWS Bedrock is heading

More frontier models added (Claude 5, Llama 4, GPT-class via OpenAI partnership rumored) through 2026.
Expanded Bedrock Agents and Knowledge Bases for managed RAG and agentic workflows.
Wider regional rollout including new EU sovereign and Asia regions.
Tighter integration with Amazon Q, SageMaker and Nova multimodal generation.

All supported LLM providers

Same endpoint, same SDK — just change the model and the BYOK header.

Ship on AWS Bedrock in 60 seconds

Free tier — bring your own AWS Bedrock key, zero markup, OpenAI-compatible endpoint.

Get your VerticalAPI key →