GPT-6 Luna
OpenAI — GPT-6 Luna
Model details
GPT-6 Luna is designed for repeatable work at scale. It can summarize documents, extract information, and answer focused questions. It accepts text and images and returns text, including code. You can adjust its reasoning effort to fit your workload.
Model launch date: September 22, 2026
EOL no sooner than: September 22, 2027
Legacy period: at least 6 months
Model lifecycle policy: Model lifecycle
Model EOL date: N/A
End User License Agreements and Terms of Use: View
Model lifecycle: Active
Context window: 1,050,000 tokens
Marketplace product ID:
prod-fiwlckcpwkwli
| Input modalities | Output modalities |
|---|---|
 Audio | Embedding |
 Image | Image |
| Speech | Speech |
| Text | Text |
| Video | Video |
Endpoints and APIs supported
These tables show the endpoints and APIs that GPT-6 Luna supports. See APIs supported by Amazon Bedrock and Endpoints supported by Amazon Bedrock.
Endpoint support
| Endpoint | Supported |
|---|---|
bedrock-runtime |
|
bedrock-mantle |
|
APIs supported on the bedrock-runtime endpoint
| Messages | Responses | Chat Completions | Converse | Invoke |
|---|---|---|---|---|
|
|
|
|
|
APIs supported on the bedrock-mantle endpoint
| Messages | Responses | Chat Completions | Converse | Invoke |
|---|---|---|---|---|
|
|
|
|
|
Note
On bedrock-mantle, both APIs use the /openai/v1 base path. Do not use /v1. Use either API with this model:
For Responses, use
/openai/v1/responses.For Chat Completions, use
/openai/v1/chat/completions.
Tip
For new applications, use the bedrock-runtime endpoint when possible. See Endpoints supported by Amazon Bedrock for details.
Capabilities and features
Amazon Bedrock features
Features supported on the bedrock-runtime endpoint
| Supported | Not Supported |
|---|---|
|
Features supported on the bedrock-mantle endpoint
| Supported | Not Supported |
|---|---|
|
Reasoning effort
Set reasoning effort to none, low,
medium, high, xhigh, or max.
The default is medium.
Pricing
All prices are in USD per 1 million tokens for the Standard tier.
Mantle in-Region and US geographic cross-Region inference include a 10% premium. The base rates are OpenAI first-party Standard rates. Global cross-Region inference uses those rates with no premium. The prices below already include any premium.
Priority and Flex tiers are not supported for these inference options.
Commercial Regions — short context (272K input tokens or fewer)
| Inference option | Input | Input — cache write | Input — cache read | Output |
|---|---|---|---|---|
| Mantle in-Region | $0.11 | $0.1375 | $0.011 | $0.55 |
| US Geo CRIS | $0.11 | $0.1375 | $0.011 | $0.55 |
| Global CRIS | $0.10 | $0.125 | $0.01 | $0.50 |
Commercial Regions — long context (more than 272K input tokens)
| Inference option | Input | Input — cache write | Input — cache read | Output |
|---|---|---|---|---|
| Mantle in-Region | $0.22 | $0.275 | $0.022 | $0.825 |
| US Geo CRIS | $0.22 | $0.275 | $0.022 | $0.825 |
| Global CRIS | $0.20 | $0.25 | $0.02 | $0.75 |
Long-context rates apply to the full request when input exceeds 272,000 tokens.
Note
Prices are subject to change. For current prices, see
Amazon Bedrock pricing
Call the model
Use these model IDs and endpoint URLs to call the model. See APIs supported and Endpoints supported.
| Endpoint | Model ID | In-Region endpoint URL | Geo inference ID | Global inference ID |
|---|---|---|---|---|
bedrock-mantle |
openai.gpt-6-luna |
https://bedrock-mantle.{region}.api.aws/openai/v1 |
Not supported | Not supported |
bedrock-runtime |
openai.gpt-6-luna |
Not supported | us.openai.gpt-6-luna |
global.openai.gpt-6-luna |
On bedrock-runtime:
Use this base URL:
https://bedrock-runtime.{region}.amazonaws.com/openai/v1.Set the model ID to
us.openai.gpt-6-lunaorglobal.openai.gpt-6-luna.Choose a profile that is available in your source Region.
You cannot use the base model ID for in-Region calls on this endpoint.
On bedrock-mantle, use openai.gpt-6-luna with the
/openai/v1 base path. Use the US East (N. Virginia) (us-east-1) Region for this model on this endpoint.
Service tiers
This model supports only the Standard tier in Amazon Bedrock. You pay per token with no commitment. Set "service_tier": "default" or omit the field. Priority, Flex, and Reserved are not supported. For more information, see service tiers.
| Standard | Priority | Flex | Reserved |
|---|---|---|---|
|
|
|
|
Supported Regions
Each endpoint supports a different set of Regions. See Regional availability by models.
The bedrock-mantle endpoint
| Region | In-Region | Geo | Global |
|---|---|---|---|
us-east-1 (US East (N. Virginia)) | Supported | Not supported | Not supported |
The bedrock-runtime endpoint
| Source Region | In-Region | US Geo CRIS | Global CRIS |
|---|---|---|---|
us-east-1 | Not supported | Supported | Supported |
us-east-2 | Not supported | Supported | Supported |
us-west-1 | Not supported | Supported | Supported |
us-west-2 | Not supported | Supported | Supported |
ca-central-1 | Not supported | Supported | Supported |
ca-west-1 | Not supported | Supported | Supported |
eu-central-1 | Not supported | Not supported | Supported |
eu-central-2 | Not supported | Not supported | Supported |
eu-north-1 | Not supported | Not supported | Supported |
eu-south-1 | Not supported | Not supported | Supported |
eu-south-2 | Not supported | Not supported | Supported |
eu-west-1 | Not supported | Not supported | Supported |
eu-west-2 | Not supported | Not supported | Supported |
eu-west-3 | Not supported | Not supported | Supported |
ap-east-2 | Not supported | Not supported | Supported |
ap-northeast-1 | Not supported | Not supported | Supported |
ap-northeast-2 | Not supported | Not supported | Supported |
ap-northeast-3 | Not supported | Not supported | Supported |
ap-south-1 | Not supported | Not supported | Supported |
ap-south-2 | Not supported | Not supported | Supported |
ap-southeast-1 | Not supported | Not supported | Supported |
ap-southeast-2 | Not supported | Not supported | Supported |
ap-southeast-3 | Not supported | Not supported | Supported |
ap-southeast-4 | Not supported | Not supported | Supported |
ap-southeast-5 | Not supported | Not supported | Supported |
ap-southeast-6 | Not supported | Not supported | Supported |
ap-southeast-7 | Not supported | Not supported | Supported |
il-central-1 | Not supported | Not supported | Supported |
af-south-1 | Not supported | Not supported | Supported |
sa-east-1 | Not supported | Not supported | Supported |
mx-central-1 | Not supported | Not supported | Supported |
Quotas and limits
Quotas vary by account and Region. To review your quotas or request an increase, see Quotas for Amazon Bedrock and Request an increase to a quota.
On bedrock-runtime, the quota counts output tokens at a 10-to-1 rate.
Each output token uses 10 tokens of quota.
Sample code
Step 1 - Create an AWS account: If you already have an AWS account, skip this step. Otherwise, sign up for an AWS account
Step 2 - Create an API key: Open the Amazon Bedrock console
Step 3 - Install the SDK: You need Python to run these examples. Install the OpenAI SDK with the command below. Both APIs use this SDK.
Step 4 - Set environment variables
Choose the tab for your endpoint. Set your API key and the base URL shown there.
Note
On bedrock-runtime, use a cross-Region inference profile as the model: us.openai.gpt-6-luna or global.openai.gpt-6-luna. This model does not support in-Region calls on this endpoint.
Your IAM identity needs bedrock:InvokeModel permission for both resources:
The inference profile.
Your AWS account's default project:
arn:aws:bedrock:{region}:{account-id}:project/default.
Step 5 - Send your first request
Save one of the examples below as bedrock-first-request.py.
bedrock-mantle
Use the settings from Step 4 - Set environment variables. Choose the bedrock-mantle tab.
bedrock-runtime: OpenAI SDK
Use the settings from Step 4 - Set environment variables. Choose the bedrock-runtime tab. Use either API. Both examples use the US system inference profile.
