What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
To check GPT-6 availability, first identify the exact model card and endpoint, then verify both the source Region and your account’s live entitlement. For the GPT-6 Astra entry currently documented by AWS, bedrock-mantle uses openai.gpt-6-astra only in us-west-2 (Oregon), while bedrock-runtime uses the cross-Region inference profiles us.openai.gpt-6-astra and global.openai.gpt-6-astra. A model appearing in AWS documentation does not prove that your account is authorized to invoke it.
Use the AWS model card, the Bedrock console, and the GetFoundationModelAvailability API together. Availability, agreements, entitlements, authorization, quotas, and regional routing can change, so check the live account immediately before deploying.
As an Amazon Associate I earn from qualifying purchases.
Is GPT-6 available in my AWS region?
There is no single yes-or-no answer based only on the Region selector in the AWS console. “Available” can mean that a model can be invoked from your source Region, that processing stays inside that Region, or that AWS accepts the request and routes it to another Region. Those are different conditions.
Recommended Free Tools
The current AWS documentation for GPT-6 Astra distinguishes two Bedrock interfaces:
#1 Best Overall
| Bedrock interface | Model or profile ID | Regional behavior documented by AWS | Endpoint |
|---|---|---|---|
bedrock-mantle |
openai.gpt-6-astra |
Listed only in us-west-2 (Oregon) |
https://bedrock-mantle.us-west-2.api.aws/openai/v1 |
bedrock-runtime |
us.openai.gpt-6-astra (Geo) or global.openai.gpt-6-astra (Global) |
Invoked through cross-Region inference profiles; the model card does not list In-Region use | https://bedrock-runtime.{region}.amazonaws.com/openai/v1 |
These identifiers and locations come from the AWS GPT-6 Astra model card. Do not assume another GPT-6 variant has the same IDs, Regions, or endpoint rules.
What “In-Region,” “Geo Cross-Region,” and “Global Cross-Region” mean
| Routing label | Where a request may be processed | Residency implication |
|---|---|---|
| In-Region | Inside the single AWS Region specified for the request | Provides the narrowest geographic boundary, subject to the service’s documented terms |
| Geo Cross-Region | Within a defined geography, such as the United States or European Union | Can leave the source Region but remains within the defined geography |
| Global Cross-Region | Any supported commercial Region worldwide | Does not provide a geographic residency restriction tied to the source Region |
A source Region is therefore not necessarily the processing Region. Choose the routing mode according to your data-residency commitment, compliance requirements, latency expectations, scale, and cost plan. AWS describes these distinctions in its regional availability by models documentation.
Start with the exact GPT-6 model card
Before checking a Region, open the current Amazon Bedrock model catalog and select the exact GPT-6 model you intend to use. The documentation available for this guide identifies GPT-6 Astra; a future GPT-6 release could have a different model ID, service interface, inference profile, or regional table.
Quick wins for a faster PC:
Clear out junk files and repair common Windows errorsFree Scan →Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →- Open the model card. Confirm the model name, supported Bedrock interface, model ID, inference profile IDs, routing types, and service tier.
- Record the invocation path. Decide whether your application will call
bedrock-mantleorbedrock-runtime. Do not substitute a base model ID for an inference-profile ID. - Read the regional table for that exact entry. A Region listed for one endpoint does not automatically apply to the other endpoint.
- Check the date of the information. AWS can change model availability and entitlements without changing your application code’s assumptions.
For GPT-6 Astra, the model card lists the standard service tier. It does not list Priority or Flex support, so do not configure those tiers unless a newer card explicitly adds them.
How to check Bedrock model access for your account
AWS says access to Bedrock foundation models is enabled by default, but account-level state still determines whether a particular model can be used. A public regional table cannot establish your account’s current agreement, authorization, entitlement, or quota status.
Use the Bedrock console
- Sign in to the AWS account and select the Region from which your application will submit requests.
- Open the Amazon Bedrock console and go to the model catalog.
- Search for the exact GPT-6 model name, then open its model details.
- Review the displayed model ID or inference profile ID and the supported endpoint.
- Open the model in a Bedrock playground, if the console offers that action.
- Run a minimal, non-sensitive test request. A successful playground request confirms that the selected console Region, model path, and account permissions align for that test; it does not prove that every other Region, profile, workload, or quota is available.
If the model is visible but the playground cannot run it, treat that as an account or configuration issue rather than proof that AWS has removed the model from the Region. Continue with the API availability check and the failure-specific diagnostics below.
Use GetFoundationModelAvailability for an account-specific result
The authoritative programmatic check is AWS’s GetFoundationModelAvailability API. Call it for the exact model ID and Region relevant to your deployment, then inspect every returned availability field:
regionAvailability— whether the model or profile is available in the queried Region.entitlementAvailability— whether the account has an available entitlement.authorizationStatus— whether the account is authorized.agreementAvailability— whether the required agreement state is available.
AWS defines authorization values of AUTHORIZED and NOT_AUTHORIZED; entitlement values of AVAILABLE and NOT_AVAILABLE; and regional values of AVAILABLE and NOT_AVAILABLE. Interpret the fields together rather than treating one failed field as a regional verdict. The API reference is at GetFoundationModelAvailability.
For example, a model can be listed publicly while your account reports NOT_AUTHORIZED. Conversely, an authorization or quota error does not prove that the model is absent from the Region. Record the model ID, queried Region, returned fields, timestamp, and request identity when troubleshooting.
What each check proves
| Evidence | What it establishes | What it does not establish |
|---|---|---|
| AWS model card | The documented IDs, endpoints, routing types, service tier, and listed source Regions | That your account has authorization or an entitlement |
| Model catalog listing | That AWS exposes a model entry to the selected console context | That the production role can invoke it successfully |
| Playground test | That one account, Region, identity, and request path worked at test time | That another Region, workload, profile, or quota will work |
GetFoundationModelAvailability |
The returned agreement, authorization, entitlement, and regional state for the queried model and Region | That a different model ID or source Region has the same state |
| Discovery API result | Which model or profile IDs the selected endpoint exposes | That every exposed ID is authorized for every operation |
Which model ID should you invoke?
The endpoint determines the identifier you pass to your client:
| If you use | Use this GPT-6 Astra identifier | Important qualification |
|---|---|---|
bedrock-mantle |
openai.gpt-6-astra |
Documented only in us-west-2; this is an In-Region service entry in the current card |
bedrock-runtime with geographic routing |
us.openai.gpt-6-astra |
Geo Cross-Region profile; processing can occur in supported US Regions rather than necessarily the source Region |
bedrock-runtime with worldwide routing |
global.openai.gpt-6-astra |
Global Cross-Region profile; processing can occur in supported commercial Regions worldwide |
Do not type a friendly display name such as “GPT-6” into an SDK call when AWS expects an exact model or profile ID. Copy the identifier shown in the current model card or returned by discovery APIs.
Discover IDs instead of hard-coding assumptions
AWS recommends using ListFoundationModels and ListInferenceProfiles to discover what the bedrock-runtime endpoint exposes to your account. The relevant instructions are in Get list of models.
For bedrock-mantle, use its OpenAI-compatible /models endpoint. AWS notes that discovery occurs through /models even though inference uses the /openai/v1 path. A discovery result is useful because it reflects what your selected endpoint and credentials can see; still verify authorization and entitlement before treating a listed ID as invokable.
Where can bedrock-runtime GPT-6 requests originate?
The current GPT-6 Astra card lists Geo and Global cross-Region support in the US and Canada source Regions us-east-1, us-east-2, us-west-1, us-west-2, and ca-central-1. It also lists Global support in specified European and Asia-Pacific Regions and in São Paulo. The card’s table is the authority for the complete, current set; do not infer that every Bedrock Region is supported.
When evaluating a source Region, check all of these dimensions:
Do these 3 things before closing this tab:
1Repair Windows errors before they cause bigger problems2Scan for outdated or missing drivers - takes under a minute3Clear out junk files and repair common Windows errors- Source-Region listing: Is the Region listed for the exact Geo or Global profile?
- Profile type: Does your workload permit Geo routing, Global routing, or only In-Region processing?
- Processing location: Could the request leave the source Region or geography?
- Account state: Does
GetFoundationModelAvailabilityreturn available agreement, authorization, entitlement, and regional status? - Operational limits: Are your quotas and IAM permissions sufficient for the intended request volume?
Why can’t I invoke GPT-6 in Bedrock?
Invocation failures have several causes that look similar from an application’s perspective. Work through them in this order.
1. The endpoint and ID do not match
Using openai.gpt-6-astra against bedrock-runtime, or using a runtime profile against bedrock-mantle, can produce a rejection even when the model is documented for your account. Compare the client’s endpoint, Region, and ID with the model-card table.
2. You selected a source Region that is not listed
A cross-Region profile is not automatically callable from every AWS Region. Check the current regional compatibility table for the exact profile and source Region. For bedrock-mantle, the documented GPT-6 Astra location is us-west-2; selecting another Region will not create an In-Region endpoint.
3. The account is not authorized or lacks an entitlement
Inspect authorizationStatus and entitlementAvailability in GetFoundationModelAvailability. A public listing can coexist with NOT_AUTHORIZED or NOT_AVAILABLE for your account.
Crashes, No Sound, or Screen Glitches?
Random freezes, missing sound and display glitches usually trace back to one bad driver. Find and replace yours safely.Free scan · under a minuteWindows Errors? Fix Them Before They Spread
Repair common Windows errors and clear accumulated junk for a smoother, more stable PC - no reinstall needed.Free scan · no reinstall4. The agreement state is incomplete
Check agreementAvailability. If the required agreement is unavailable, resolve the account’s Bedrock agreement or access workflow before changing application code.
5. IAM permissions block the request
Confirm that the calling identity is allowed to discover models, use the selected Bedrock endpoint, and invoke the specific model or inference profile. Test with the same role, account, and Region used by the application; a successful test under an administrator identity does not validate a production role.
6. Quotas or service tier rules apply
AWS states that accounts have default quotas and that Region and other account factors can affect them. A quota or throttling error is an operational limit, not evidence that GPT-6 is unavailable in the Region. Check the account’s quota values and the service tier supported by the model card before raising limits or redesigning routing.
7. Residency requirements conflict with the chosen profile
If policy requires processing to remain in one Region, a Geo or Global profile may be unacceptable even when invocation succeeds. Switch to a documented In-Region option only if the exact model card provides one; do not describe cross-Region access as In-Region processing.
The Tool Desk
Outbyte PC Repair FREERepair Windows errors before they cause bigger problemsFix Now →Outbyte Driver Updater FREEScan for outdated or missing drivers - takes under a minuteDriver Scan →How to choose between the documented GPT-6 paths
| Decision factor | bedrock-mantle |
bedrock-runtime Geo |
bedrock-runtime Global |
|---|---|---|---|
| API compatibility | OpenAI-compatible Mantle endpoint | Bedrock Runtime with us.openai.gpt-6-astra |
Bedrock Runtime with global.openai.gpt-6-astra |
| Documented GPT-6 Astra ID | openai.gpt-6-astra |
us.openai.gpt-6-astra |
global.openai.gpt-6-astra |
| Source-Region rule | Listed only in us-west-2 |
Use only a source Region listed for the Geo profile | Use only a source Region listed for the Global profile |
| Potential processing area | Single documented Region | Defined geography | Supported commercial Regions worldwide |
| Best fit | Workloads that can use the Mantle API and operate from Oregon | Workloads needing a US geographic boundary rather than one exact Region | Workloads prioritizing broad routing flexibility and accepting worldwide processing |
The right choice depends on endpoint compatibility, source-Region support, residency rules, account authorization, and quotas—not merely on which option appears in a catalog search.
Cost, quota, and service-tier considerations
The AWS materials used for this check do not provide a single GPT-6 price or a universal quota number. Treat cost as a deployment decision rather than as evidence of availability. Compare the endpoint and routing mode you are considering with your workload’s scale and residency requirements, then consult the current AWS commercial terms and account quota view for the figures that apply to your account.
No hardware, accessory, book, or manual is required to perform the check. You need AWS credentials and permissions suitable for the console, discovery calls, availability API, and eventual invocation. Those credentials are account prerequisites, not a separate product purchase.
The GPT-6 Astra card lists the standard service tier and does not list Priority or Flex. A request that specifies an unsupported tier can fail even when the model and Region are otherwise correct. Quotas are also account- and Region-dependent, so validate expected request volume separately from basic model access.
Free tools Windows power users keep installed
One-click scans. No signup required.
Examples of defensible availability conclusions
- “Mantle is available for this test account in Oregon.” This conclusion is defensible only when the account check and a test using
openai.gpt-6-astraat theus-west-2Mantle endpoint succeed. - “Runtime Geo can be invoked from this listed US source Region.” Confirm the Region appears in the current Geo table, use
us.openai.gpt-6-astra, and verify account fields. Do not claim that processing stayed in that source Region. - “Runtime Global is callable from this listed source Region.” Confirm the Global profile and account state. The request may be processed in a supported commercial Region worldwide, so this statement says nothing about single-Region residency.
- “The model is documented but not enabled for this account.” Use this wording when the model card lists the path but the account returns
NOT_AUTHORIZED,NOT_AVAILABLE, or an unavailable agreement state.
A repeatable pre-deployment checklist
- Write down the exact GPT-6 variant and version shown in the current AWS model card.
- Choose
bedrock-mantleorbedrock-runtimebefore selecting a model ID. - For Mantle, confirm
openai.gpt-6-astraandus-west-2; for Runtime, select the documented Geo or Global profile. - Compare your source Region with the model card’s current regional table.
- Decide whether In-Region, Geo Cross-Region, or Global Cross-Region processing is acceptable for the data.
- Use the console model catalog and a playground test with the same account and Region as production.
- Call
GetFoundationModelAvailabilityand save the four account-specific availability fields. - Discover the endpoint-visible ID with
ListFoundationModels,ListInferenceProfiles, or Mantle’s/modelsendpoint. - Validate IAM permissions, quotas, and the supported service tier.
- Log the Region, profile ID, account, timestamp, and routing mode so a later AWS availability change can be distinguished from an application regression.
What to recheck when availability changes
Availability is not a permanent property of a model name. Recheck the model card and regional compatibility table after a new model variant, endpoint migration, account change, Region change, quota change, or failed deployment. The current authoritative references are AWS’s GPT-6 Astra model card, Using models with Bedrock, GetFoundationModelAvailability API reference, model-discovery guide, and regional compatibility table. The information in this article was checked against those materials on October 13, 2026 UTC; AWS documentation and account entitlements remain the final authority.
Best Value
Frequently Asked Questions
Is GPT-6 available in my AWS region?
That depends on the exact GPT-6 variant, endpoint, source Region, and account. For GPT-6 Astra, Mantle is documented only in us-west-2. Bedrock Runtime uses Geo or Global cross-Region profiles from Regions listed in the current model-card table, not a single In-Region GPT-6 endpoint. Check the table and then confirm your account with GetFoundationModelAvailability.
How do I check Bedrock model access for my account?
Open the model in the Bedrock catalog and playground using the same account and Region as production, then call GetFoundationModelAvailability for the exact model ID and Region. Inspect regionAvailability, entitlementAvailability, authorizationStatus, and agreementAvailability together.
Why can’t I invoke GPT-6 in Bedrock?
Check for an endpoint and ID mismatch, an unlisted source Region, unavailable authorization or entitlement, an incomplete agreement, missing IAM permissions, quota limits, or an unsupported service tier. A model’s public listing does not prove that your account can invoke it, and a quota or authorization error does not prove that the model is absent from the Region.
Does a Bedrock Runtime source Region guarantee that GPT-6 runs there?
No. The documented Runtime profiles are cross-Region. Geo routing can process a request elsewhere within its defined geography, while Global routing can use supported commercial Regions worldwide. A source Region identifies where you submit the request, not necessarily where processing occurs.
Can I use the Mantle model ID with Bedrock Runtime?
Do not assume so. The documented Mantle ID is openai.gpt-6-astra, while Runtime uses us.openai.gpt-6-astra or global.openai.gpt-6-astra. Match the identifier to the endpoint shown in the current model card and discovery results.
Do I need to buy hardware or a special product to test GPT-6 in Bedrock?
No physical product is required. The documented workflow uses an AWS account, suitable credentials and IAM permissions, the Bedrock console or APIs, and a supported endpoint. AWS’s GPT-6 Astra card lists the standard service tier rather than Priority or Flex.
The Bottom Line
Check GPT-6 Astra in three layers: the exact AWS model card, the source-Region and routing table, and your account’s live availability response. Use openai.gpt-6-astra only with the documented bedrock-mantle endpoint in us-west-2; use us.openai.gpt-6-astra or global.openai.gpt-6-astra with the listed bedrock-runtime cross-Region paths. Then verify agreement, authorization, entitlement, IAM permissions, quotas, and residency requirements before treating the deployment as ready.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
