Amazon Bedrock Mantle is an OpenAI-compatible API for Amazon Bedrock. It allows developers to use models such as GPT-OSS 120B with familiar OpenAI APIs, including Chat Completions and Responses.

Instead of sending requests to OpenAI, your application connects to AWS and uses:

openai.gpt-oss-120b

This makes it easy to adapt applications already using the OpenAI SDK.

Availability in Europe

Bedrock Mantle is not currently available in the AWS Paris region. A good option for applications in France is therefore Frankfurt, Germany (eu-central-1), where both Mantle and GPT-OSS 120B are available.

The endpoint is:

https://bedrock-mantle.eu-central-1.api.aws/v1

Pricing

GPT-OSS 120B is charged based on the number of input and output tokens processed.

AWS pricing is region-dependent. AWS’s current public pricing page does not clearly publish a separate Frankfurt rate for GPT-OSS 120B, so developers should check the AWS pricing page or console before deployment.

As a reference, AWS currently lists GPT-OSS 120B in Sydney at:

• $0.1545 / 1M input tokens
• $0.618 / 1M output tokens

For example, 100,000 input tokens and 20,000 output tokens would cost about $0.028 at those rates.

Using it

Developers can use the standard OpenAI Python SDK:

from openai import OpenAI

client = OpenAI(
    base_url="https://bedrock-mantle.eu-central-1.api.aws/v1",
    api_key="YOUR_BEDROCK_API_KEY"
)

response = client.responses.create(
    model="openai.gpt-oss-120b",
    input="Explain neural networks simply."
)

print(response.output_text)

Bedrock Mantle can be useful for AI assistants, coding tools, document analysis, agents and content generation, particularly for developers who already use the OpenAI API format.

Official website: Amazon Bedrock Amazon Bedrock official website