# GPT-6.1 Sol
##  OpenAI — GPT-6.1 Sol
## Model Details
OpenAI GPT-6.1 Sol brings advanced capabilities to workloads where both performance and cost matter. GPT-6.1 Sol helps agents investigate codebases and iterate on solutions. Agents can also use it to understand complex documents and complete business and computer use workflows across multiple steps.
+ **Model launch date:** September 29, 2026
+ **Model lifecycle policy:** [OpenAI model deprecation notice periods](https://developers.openai.com/api/docs/deprecations#model-deprecation-notice-periods). This model follows OpenAI first-party lifecycle terms, with at least 6 months of deprecation notice for generally available models, unless safety or compliance concerns require a faster timeline.
+ **Model EOL date:** Not announced.
+ **End User License Agreements and Terms of Use:** [OpenAI models on Amazon Bedrock terms](https://aws.amazon.com/legal/bedrock/third-party-models/#amsc13--xttmgl)
+ **Model lifecycle:** Active
+ **Context window:** 1M tokens
+ **Max output tokens:** 131,072 tokens
+ **Marketplace product ID:** `prod-qco655ut2vn54`
| **Input Modalities** | **Output Modalities** |
| --- | --- |
|  Audio |  Embedding |
|  Image |  Image |
|  Speech |  Speech |
|  Text |  Text |
|  Video |  Video |
## Endpoints and APIs supported
The following tables show which endpoints and APIs GPT-6.1 Sol supports. For more information, see [APIs supported by Amazon Bedrock](apis.md) and [Endpoints supported by Amazon Bedrock](endpoints.md).
**Endpoint support**
| **Endpoint** | **Supported** |
| --- | --- |
| bedrock-runtime |  |
| bedrock-mantle |  |
**APIs supported on `bedrock-runtime`**
| **Messages** | **Responses** | **Chat Completions** | **Converse** | **Invoke** |
| --- | --- | --- | --- | --- |
|  |  |  |  |  |
**APIs supported on `bedrock-mantle`**
| **Messages** | **Responses** | **Chat Completions** | **Converse** | **Invoke** |
| --- | --- | --- | --- | --- |
|  |  |  |  |  |
**Note**
On `bedrock-mantle`, both APIs use the `/openai/v1` base path, not `/v1`. Use either API with this model:
For Responses, use `/openai/v1/responses`.
For Chat Completions, use `/openai/v1/chat/completions`.
## Capabilities and Features
***Bedrock Features***
**Features supported using `bedrock-runtime`**
| **Supported** | **Not Supported** |
| --- | --- |
| +  [Projects (default project only)](projects.html)
+  [Invocation logs](model-invocation-logging.html)
+  [Response streaming](/bedrock/latest/APIReference/API_runtime_InvokeModelWithResponseStream.html)
+  [Abuse detection](abuse-detection.html)
+  [Guardrails](guardrails.html) (Chat Completions, InvokeModel, and Converse APIs; not supported with Responses)
+  [Application inference profiles](cost-mgmt-application-inference-profiles.html) (InvokeModel and Converse APIs only; not supported with Responses or Chat Completions)
+  [Implicit Prompt Caching](#model-card-openai-gpt-6-1-sol-prompt-caching) (Responses, Chat Completions, InvokeModel, and Converse APIs)
+  [Explicit Prompt Caching](#model-card-openai-gpt-6-1-sol-prompt-caching) (Responses, Chat Completions, and InvokeModel APIs; not supported with Converse)
+  [Structured outputs](#model-card-openai-gpt-6-1-sol-structured-output) (Responses, Chat Completions, InvokeModel, and Converse APIs) | +  [Server-side system tools](tool-use.html)
+  [Intelligent prompt routing](prompt-routing.html)
+  [Count tokens](count-tokens.html)
+  [Knowledge Bases](knowledge-base.html)
+  [Prompt optimization](prompt-management-optimize.html) |
**Features supported using `bedrock-mantle`**
| **Supported** | **Not Supported** |
| --- | --- |
| +  [Projects](projects.html)
+  [Response streaming](inference-responses-api.md#bedrock-mantle-responses-streaming)
+  [Implicit Prompt Caching](#model-card-openai-gpt-6-1-sol-prompt-caching) (Responses and Chat Completions APIs)
+  [Explicit Prompt Caching](#model-card-openai-gpt-6-1-sol-prompt-caching) (Responses and Chat Completions APIs)
+  [Structured outputs](#model-card-openai-gpt-6-1-sol-structured-output) (Responses and Chat Completions APIs) | +  [Guardrails](guardrails.html)
+  [Invocation logs](model-invocation-logging.html)
+  [Application inference profiles](cost-mgmt-application-inference-profiles.html)
+  [Server-side system tools](tool-use.html)
+  [Intelligent prompt routing](prompt-routing.html)
+  [Count tokens](count-tokens.html)
+  [Knowledge Bases](knowledge-base.html)
+  [Prompt optimization](prompt-management-optimize.html) |
### Prompt caching
GPT-6.1 Sol supports both [Implicit Prompt Caching](prompt-caching.html#prompt-caching-implicit) and [Explicit Prompt Caching](prompt-caching.html#prompt-caching-explicit). API support differs by endpoint:
+ Responses and Chat Completions support both caching types on the `bedrock-runtime` and `bedrock-mantle` endpoints.
+ InvokeModel supports both caching types on the `bedrock-runtime` endpoint.
+ Converse supports implicit caching on the `bedrock-runtime` endpoint. The native `cachePoint` field isn't supported for explicit caching with this model.
For explicit caching, set `prompt_cache_options.mode` to `explicit`. Set `prompt_cache_options.ttl` to `30m`, the only supported TTL and the default. In Responses requests, add `"prompt_cache_breakpoint": {"mode": "explicit"}` to an `input_text` content block. In Chat Completions and InvokeModel requests, add it to a `text` content part.
A cacheable prompt prefix must contain at least 1,024 tokens. You can include multiple explicit breakpoints, and each request can create up to four cache writes. Check the cache usage fields in the response to determine whether tokens were written to or read from the cache.
### JSON Schema output
Use JSON Schema to set the format of the model response for streaming and non-streaming requests. Chat Completions and Responses support JSON Schema output on both the `bedrock-runtime` and `bedrock-mantle` endpoints. InvokeModel, InvokeModelWithResponseStream, Converse, and ConverseStream support JSON Schema output on the `bedrock-runtime` endpoint. Choose a model ID or inference profile ID from [Programmatic access](#model-card-openai-gpt-6-1-sol-programmatic-access).
+ For Chat Completions, set `response_format.type` to `json_schema`. Put `name`, `schema`, and `strict: true` in `response_format.json_schema`.
+ For Responses, set `text.format.type` to `json_schema`. Put `name`, `schema`, and `strict: true` in `text.format`.
+ For InvokeModel and InvokeModelWithResponseStream, use the same `response_format` fields as Chat Completions.
+ For Converse and ConverseStream, set `outputConfig.textFormat.type` to `json_schema`. In `outputConfig.textFormat.structure.jsonSchema`, set `name` and `schema`. Encode the schema as a JSON string. Then set `additionalModelRequestFields.text.format.strict` to `true`.
Use an object schema. Put all fields in `required`. Set `additionalProperties` to `false`. Validate your schema before you send it. Check for refusals or incomplete responses before you parse the output.
**Important**
For JSON Schema output with Converse and ConverseStream, you must set `additionalModelRequestFields.text.format.strict` to `true`, in addition to specifying the schema in `outputConfig`. Include the following field in your request.
```
{
"additionalModelRequestFields": {
"text": {
"format": { "strict": true }
}
}
}
```
## Pricing
All prices are in USD per 1 million tokens. The following tables list Standard and Ultrafast prices. Global CRIS rates match [OpenAI first-party pricing](https://developers.openai.com/api/docs/pricing) for the same service tier.
In-Region access on `bedrock-mantle` costs 10% more than Global CRIS. US geographic cross-Region inference (US CRIS) on `bedrock-runtime` has the same premium. This applies to both service tiers. The tables include this premium.
GPT-6.1 Sol supports both implicit and explicit prompt caching. Cache-write tokens are billed at 1.25× the uncached input-token rate, and cache-read tokens are billed at 0.05× the uncached input-token rate. For configuration and API support, see [Prompt caching](#model-card-openai-gpt-6-1-sol-prompt-caching).
Long-context rates apply to the full request when input exceeds 272,000 tokens.
*Priority and Flex tiers are not supported for this model.*
### Standard — Commercial Regions, short context (272K input tokens or fewer)
| **Inference option** | **Input** | **Input — cache write** | **Input — cache read** | **Output** |
| --- | --- | --- | --- | --- |
| Regional (bedrock-mantle in US East (N. Virginia)) | $2.20 | $2.75 | $0.11 | $11.00 |
| US CRIS (bedrock-runtime) | $2.20 | $2.75 | $0.11 | $11.00 |
| Global CRIS (bedrock-runtime) | $2.00 | $2.50 | $0.10 | $10.00 |
### Standard — Commercial Regions, long context (more than 272K input tokens)
| **Inference option** | **Input** | **Input — cache write** | **Input — cache read** | **Output** |
| --- | --- | --- | --- | --- |
| Regional (bedrock-mantle in US East (N. Virginia)) | $4.40 | $5.50 | $0.22 | $16.50 |
| US CRIS (bedrock-runtime) | $4.40 | $5.50 | $0.22 | $16.50 |
| Global CRIS (bedrock-runtime) | $4.00 | $5.00 | $0.20 | $15.00 |
Ultrafast costs six times as much as Standard. Global CRIS rates match [OpenAI first-party Ultrafast pricing](https://developers.openai.com/api/docs/pricing?latest-pricing=ultrafast). On `bedrock-runtime`, use Ultrafast through US CRIS or Global CRIS. On `bedrock-mantle`, use Ultrafast in `us-east-1`. Short-context prices apply to requests with up to 272,000 input tokens. Above that limit, long-context prices apply to the full request.
### Ultrafast — Commercial Regions, short context (272K input tokens or fewer)
| **Inference option** | **Input** | **Input — 30m cache write** | **Input — cache read** | **Output** |
| --- | --- | --- | --- | --- |
| In-Region (bedrock-mantle in us-east-1) | $13.20 | $16.50 | $0.66 | $66.00 |
| US CRIS (bedrock-runtime) | $13.20 | $16.50 | $0.66 | $66.00 |
| Global CRIS (bedrock-runtime) | $12.00 | $15.00 | $0.60 | $60.00 |
### Ultrafast — Commercial Regions, long context (more than 272K input tokens)
| **Inference option** | **Input** | **Input — 30m cache write** | **Input — cache read** | **Output** |
| --- | --- | --- | --- | --- |
| In-Region (bedrock-mantle in us-east-1) | $26.40 | $33.00 | $1.32 | $99.00 |
| US CRIS (bedrock-runtime) | $26.40 | $33.00 | $1.32 | $99.00 |
| Global CRIS (bedrock-runtime) | $24.00 | $30.00 | $1.20 | $90.00 |
## Programmatic Access
To call this model from code, use the following model IDs and endpoint URLs. For more information, see [APIs supported by Amazon Bedrock](apis.md) and [Endpoints supported by Amazon Bedrock](endpoints.md).
| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** |
| --- | --- | --- | --- | --- |
| bedrock-mantle | openai.gpt-6.1-sol | https://bedrock-mantle.us-east-1.api.aws/openai/v1 | Not supported | Not supported |
| bedrock-runtime | openai.gpt-6.1-sol | Not supported | us.openai.gpt-6.1-sol | global.openai.gpt-6.1-sol |
*For in-Region access, use `bedrock-mantle` in `us-east-1` (N. Virginia). On `bedrock-runtime`, the OpenAI-compatible base URL is `https://bedrock-runtime.{region}.amazonaws.com/openai/v1`. Use `us.openai.gpt-6.1-sol` for US geographic cross-Region inference or `global.openai.gpt-6.1-sol` for global cross-Region inference. Direct in-Region invocation is not supported on `bedrock-runtime`. Use a source Region enabled for the profile you choose; see [Route model inference requests across AWS Regions with cross-Region inference](cross-region-inference.md).*
## Service Tiers
Amazon Bedrock offers several service tiers for different workloads. **Standard** gives you pay-per-token access with no commitment. To use it, set `"service_tier": "default"` or omit the field. For more information, see [service tiers](service-tiers-inference.html).
**Ultrafast** is a premium speed tier for GPT-6.1 Sol. Use it for workloads where response speed matters most. To request it with the Responses API, set `"service_tier": "ultrafast"`.
| **Standard** | **Ultrafast** | **Priority** | **Flex** | **Reserved** |
| --- | --- | --- | --- | --- |
|  |  |  |  |  |
Use one of these routes for Ultrafast:
+ `bedrock-mantle` in `us-east-1`, with model ID `openai.gpt-6.1-sol`.
+ `bedrock-runtime`, with `us.openai.gpt-6.1-sol` for US CRIS or `global.openai.gpt-6.1-sol` for Global CRIS.
## Regional Availability
***Regional availability at a glance***
Standard and Ultrafast are available on `bedrock-mantle` in `us-east-1` (N. Virginia). Both tiers support US geographic and global inference profiles on `bedrock-runtime`. For more information, see [Regional availability by models](models-region-compatibility.md).
**Availability using the `bedrock-mantle` endpoint**
| **Region** | **In-Region** | **Geo** | **Global** |
| --- | --- | --- | --- |
| us-east-1 (N. Virginia) |  |  |  |
**Availability using the `bedrock-runtime` endpoint**
| **Scope** | **In-Region** | **Geo** | **Global** |
| --- | --- | --- | --- |
| US geographic and global inference |  |  |  |
***Geo inference details***
The destination Regions available to a geographic inference profile depend on the source Region. To retrieve the current routing configuration, call [GetInferenceProfile](https://docs.aws.amazon.com/bedrock/latest/APIReference/API_GetInferenceProfile.html) from the source Region.
**Geo: US**
Geo inference ID: `us.openai.gpt-6.1-sol`
| **Source Region** | **Destination Regions** |
| --- | --- |
| us-east-1 (N. Virginia) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) |
| us-east-2 (Ohio) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) |
| us-west-1 (N. California) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-1 (N. California), us-west-2 (Oregon) |
| us-west-2 (Oregon) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) |
| ca-central-1 (Canada) | ca-central-1 (Canada), us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) |
| ca-west-1 (Calgary) | ca-west-1 (Calgary), us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) |
## Quotas and Limits
Quotas vary by account and Region. See [Quotas for Amazon Bedrock](quotas.md) and the model's service quota settings.
The output-token burndown rate is 10: each output token consumes 10 tokens of quota.
## Sample Code
**Step 1 - AWS Account:** If you already have an AWS account, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).
**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.
**Step 3 - Get the SDK:** You must have Python installed to use this guide. Then install the OpenAI SDK.
```
python3 -m pip install openai
```
### Step 4 - Set environment variables
Set up your environment to use the API key for authentication.
------
#### [ bedrock-mantle ]
```
export OPENAI_API_KEY=""
export OPENAI_BASE_URL="https://bedrock-mantle.us-east-1.api.aws/openai/v1"
```
------
#### [ bedrock-runtime ]
```
export OPENAI_API_KEY=""
export OPENAI_BASE_URL="https://bedrock-runtime.us-east-1.amazonaws.com/openai/v1"
```
------
**Note**
On `bedrock-runtime`, set the model to `us.openai.gpt-6.1-sol` for US CRIS or `global.openai.gpt-6.1-sol` for Global CRIS. Direct in-Region invocation is not supported on this endpoint.
### Step 5 - Run your first inference request
Save the file as `bedrock-first-request.py`.
#### bedrock-mantle
Use the settings from [Step 4 - Set environment variables](#model-card-openai-gpt-6-1-sol-sample-code-environment). Choose the `bedrock-mantle` tab. The Chat tab uses the Chat Completions API.
------
#### [ Responses API ]
```
from openai import OpenAI
client = OpenAI()
response = client.responses.create(
model="openai.gpt-6.1-sol",
input="Can you explain the features of Amazon Bedrock?",
max_output_tokens=512,
)
print(response.output_text)
```
------
#### [ Chat ]
```
from openai import OpenAI
client = OpenAI()
response = client.chat.completions.create(
model="openai.gpt-6.1-sol",
messages=[{"role": "user", "content": "Can you explain the features of Amazon Bedrock?"}]
)
print(response.choices[0].message.content)
```
------
#### bedrock-runtime: OpenAI SDK
Use the settings from [Step 4 - Set environment variables](#model-card-openai-gpt-6-1-sol-sample-code-environment). Choose the `bedrock-runtime` tab. Send your request with the Responses API. The example uses the US CRIS profile. For Global CRIS, replace `us.openai.gpt-6.1-sol` with `global.openai.gpt-6.1-sol`.
------
#### [ Responses API ]
```
from openai import OpenAI
client = OpenAI()
response = client.responses.create(
model="us.openai.gpt-6.1-sol",
input="Can you explain the features of Amazon Bedrock?",
max_output_tokens=512,
)
print(response.output_text)
```
------
## Ultrafast example
Set `OPENAI_API_KEY` to your Amazon Bedrock API key. This example uses the `bedrock-mantle` endpoint in `us-east-1`.
```
export OPENAI_API_KEY=""
```
```
from openai import OpenAI
client = OpenAI(
base_url="https://bedrock-mantle.us-east-1.api.aws/openai/v1",
)
response = client.responses.create(
model="openai.gpt-6.1-sol",
service_tier="ultrafast",
input="Explain how Amazon Bedrock cross-Region inference works.",
)
print(response.output_text)
```
For `bedrock-runtime`, use the OpenAI-compatible base URL from [Programmatic Access](#model-card-openai-gpt-6-1-sol-programmatic-access). Set the model to `us.openai.gpt-6.1-sol` for US CRIS or `global.openai.gpt-6.1-sol` for Global CRIS, and keep `service_tier="ultrafast"`.