---
title: GPT-6 Sol and Luna on Amazon Bedrock (September 2026): Routing Guide
description: On September 22, 2026 AWS made GPT-6 Sol and GPT-6 Luna generally available on Bedrock. OpenAI’s internal eval found Sol makes about half as many factual mistakes as GPT-5.6 Sol. Here is the lane routing.
url: https://www.factualminds.com/blog/openai-gpt-6-sol-luna-amazon-bedrock-2026/
datePublished: 2026-09-23T00:00:00.000Z
dateModified: 2026-09-23T00:00:00.000Z
author: palaniappan-p
category: Generative AI
tags: amazon-bedrock, openai, generative-ai, finops, foundation-models, cost-optimization
---

# GPT-6 Sol and Luna on Amazon Bedrock (September 2026): Routing Guide

> On September 22, 2026 AWS made GPT-6 Sol and GPT-6 Luna generally available on Bedrock. OpenAI’s internal eval found Sol makes about half as many factual mistakes as GPT-5.6 Sol. Here is the lane routing.

> **AWS lifecycle notice (June 30, 2026)** — Amazon Bedrock Agents Classic is in maintenance for new customers after **July 30, 2026**. Net-new agent builds should use [Bedrock AgentCore](/blog/amazon-bedrock-agentcore-production/). Full matrix: [lifecycle roundup](/blog/aws-service-lifecycle-updates-june-2026/).

On **September 22, 2026**, AWS [made GPT-6 Sol and GPT-6 Luna generally available on Amazon Bedrock](https://aws.amazon.com/about-aws/whats-new/2026/09/openai-gpt-6-sol-luna-on-amazon-bedrock/). The [Machine Learning Blog post](https://aws.amazon.com/blogs/machine-learning/bring-more-intelligence-to-everyday-work-with-gpt-6-sol-and-gpt-6-luna-on-amazon-bedrock/) positions them under **GPT-6 Astra**, which was already the top of the family. Sol is the daily model for recurring complex work. Luna is the high-volume model. On an internal OpenAI factuality evaluation, **Sol makes roughly half as many mistakes as GPT-5.6 Sol**.

This is a routing event on top of the [July 30, 2026 GPT-5.6 price cuts](/blog/openai-gpt-5-6-luna-terra-bedrock-pricing-2026/). Do not flip production model IDs because the family number changed.

---

## What each model is for

AWS and OpenAI describe three GPT-6 jobs, not one default:

| Model | Job | Bedrock status |
| --- | --- | --- |
| **GPT-6 Astra** | Hardest end-to-end work: complex reasoning, coding, computer use, research, document creation | GA **September 8, 2026**. [Model card](https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-openai-gpt-6-astra.html): `openai.gpt-6-astra`. Mantle in `us-west-2`. Runtime profiles `us.openai.gpt-6-astra` and `global.openai.gpt-6-astra`. Context **1,050,000** tokens. Max output **128,000**. |
| **GPT-6 Sol** | Recurring complex tasks and software development: implement, debug, review, refactor, analyze data, multistep tools | GA **September 22, 2026**. About **half as many factual mistakes** as GPT-5.6 Sol on OpenAI’s internal factuality eval. |
| **GPT-6 Luna** | High-volume extraction, summarization, classification, and routing, with **adjustable reasoning effort** | GA **September 22, 2026**. OpenAI reports better factual reliability than the prior Luna; the announcement does not publish a second numeric eval. |

Both Sol and Luna support **up to 1M tokens** of context on Bedrock (AWS). OpenAI’s API docs list a **1,050,000**-token window for `gpt-6-sol`. Treat the Bedrock ceiling as the model card’s number once that card is in front of you — the two figures are close, and they are not the same sentence.

**Opinionated take:** new high-QPS classification, extraction, and routing go to **GPT-6 Luna**. New recurring coding and multistep tool lanes go to **GPT-6 Sol**. The hardest OpenAI jobs go to **GPT-6 Astra**. Keep [GPT-5.6 Terra and Sol](/blog/openai-gpt-5-6-luna-terra-bedrock-pricing-2026/) until a frozen-prompt bakeoff says the GPT-6 lane wins on quality **and** cost per completed task. Keep ZDR-default Claude work on [Opus 5.5](/blog/claude-opus-5-5-aws-bedrock-2026/), not on an OpenAI model, when Guardrails composition and default zero data retention are the requirement.

---

## Model IDs and endpoints

[OpenAI’s Amazon Bedrock guide](https://developers.openai.com/api/docs/guides/amazon-bedrock) (checked September 23, 2026):

| Model | Mantle (in-region) | Runtime geo | Runtime global |
| --- | --- | --- | --- |
| Sol | `openai.gpt-6-sol` in **us-east-1** | `us.openai.gpt-6-sol` | `global.openai.gpt-6-sol` |
| Luna | `openai.gpt-6-luna` in **us-east-1** | `us.openai.gpt-6-luna` | `global.openai.gpt-6-luna` |
| Astra | `openai.gpt-6-astra` in **us-west-2** | `us.openai.gpt-6-astra` | `global.openai.gpt-6-astra` |

Mantle base URL: `https://bedrock-mantle.{region}.api.aws/openai/v1`. Runtime base URL: `https://bedrock-runtime.{region}.amazonaws.com/openai/v1`, and the model name must be the inference profile — not the bare `openai.gpt-6-sol` ID. Hosted web search is Mantle-only. Background mode is Mantle-only and depends on data-retention settings. A region in the URL is not a data-residency guarantee; read the destination regions of the profile.

Context: OpenAI Python SDK, Mantle, `us-east-1`, Bedrock bearer token.

```python
import os
from openai import OpenAI

client = OpenAI(
    api_key=os.environ['AWS_BEARER_TOKEN_BEDROCK'],
    base_url='https://bedrock-mantle.us-east-1.api.aws/openai/v1',
)

response = client.responses.create(
    model='openai.gpt-6-sol',
    input='Summarize the failure modes in this migration plan and list what you could not confirm.',
)
print(response.output_text)
```

For Runtime in the same region, point `base_url` at `https://bedrock-runtime.us-east-1.amazonaws.com/openai/v1` and set `model` to `us.openai.gpt-6-sol` or `global.openai.gpt-6-sol`. The guide’s older samples still show `openai.gpt-5.6-sol` in `us-east-2`. Changing only the model ID, and leaving the region, is the failure mode.

> **What broke (pattern, not a cited client)** — A client that copied the GPT-5.6 Mantle sample, swapped in `openai.gpt-6-sol`, and left `us-east-2` in the base URL. Sol and Luna Mantle availability in the OpenAI guide is **us-east-1**. The call fails closed. Fix the region and the ID together, then confirm the live card.

---

## Prompt caching and a three-model route

AWS says Sol and Luna support **explicit prompt caching**. Mark stable instructions, tool definitions, policies, and extraction schemas so later calls process the new input. That matters when one request is classified by Luna, investigated by Sol, and escalated to Astra — each stage can reuse its own cached prefix instead of re-sending the policy pack.

Do not share one cache prefix across the three models. Cache identity is per model. A Luna checkpoint is not a Sol hit.

---

## Prices: what is published, and what is not

AWS says Sol and Luna come at **significantly lower API pricing** than GPT-5.6 and does not print a rate card in the what’s-new post or the ML blog.

OpenAI’s API docs (first-party, not a Bedrock invoice) list, per 1M tokens:

| Model | Input | Cached input | Cache write | Output |
| --- | --- | --- | --- | --- |
| GPT-6 Sol | $2.00 | $0.20 | $2.50 | $10.00 |
| GPT-6 Luna | $0.10 | $0.01 | $0.125 | $0.50 |

Prompts over **272K** input tokens are **2×** input and cache rates and **1.5×** output for the whole request. Regional processing adds **10%**. OpenAI’s Bedrock guide says commercial Bedrock pricing **matches OpenAI direct pricing**, and a region-specific Bedrock service is priced like OpenAI Regional processing. The [Bedrock pricing page](https://aws.amazon.com/bedrock/pricing/) lists GPT-6 Sol and Luna as selectable models. The static extract on September 23, 2026 did not include their dollar rows. **Confirm the selector before you re-forecast.**

GPT-6 Astra **does** publish Bedrock Standard-tier rates on its [model card](https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-openai-gpt-6-astra.html). Short context (272K input tokens or fewer), per 1M tokens:

| Inference | Input | 30-minute cache write | Cache read | Output |
| --- | --- | --- | --- | --- |
| Global CRIS | $10.00 | $12.50 | $1.00 | $50.00 |
| In-region and geo CRIS | $11.00 | $13.75 | $1.10 | $55.00 |

Long context (more than 272K input tokens) doubles input and cache rates and prices output at 1.5× ($75 global, $82.50 in-region). Priority and Flex are not supported for Astra. The card says in-region prices already include the 10% fee over OpenAI rates.

Previously published GPT-5.6 Bedrock on-demand pins ($0.22 / $1.32 Luna, $5.50 / $33 Sol) stay in our calculators. We are not replacing them with derived GPT-6 rates.

---

## Data controls AWS actually named

- IAM for access, CloudTrail for every invocation, PrivateLink for VPC-only traffic.
- Inference runs on hardware-isolated infrastructure with **zero-operator access**.
- Inference data is **not used for training**, and you do not opt in to sharing it with OpenAI.
- Classifier-flagged abuse-detection traffic is retained by AWS for **up to 30 days** and processed programmatically. Zero data retention is a request through your account team.

That is a different default from Claude Opus 5.5, where AWS says ZDR is on by default on Bedrock. If the lane’s requirement is ZDR without an account-team exception, do not move it to GPT-6 because the coding eval looks better.

---

## What to Do This Week

1. Inventory production callers still pinned to `openai.gpt-5.6-luna`, `openai.gpt-5.6-terra`, and `openai.gpt-5.6-sol`.
2. Add Sol and Luna in a non-prod account. Smoke-test Mantle in **us-east-1** and one Runtime profile (`us.` or `global.`).
3. Pick one high-volume lane and one coding lane. Freeze the prompts. Score quality, latency, and cost per completed task against the GPT-5.6 pin.
4. Turn on explicit prompt caching only for prefixes you can prove are identical across requests.
5. Block a production default change until the Bedrock price selector shows a rate you are willing to budget, and until the bakeoff passes.

---

## What This Post Doesn't Cover

- A Bedrock dollar rate for Sol or Luna copied from the pricing page. The page lists the models; this pass did not capture dollar rows. OpenAI first-party rates are labeled as such above.
- A first-party FactualMinds bakeoff. The “half as many mistakes” figure is OpenAI’s internal factuality eval, not ours.
- GPT-6 Terra. These announcements do not ship one.
- Codex or Managed Agents packaging. See the [April OpenAI-on-Bedrock note](/blog/amazon-bedrock-openai-models-codex-managed-agents/).

---

## Related reading

- [GPT-5.6 Luna and Terra Bedrock Price Cuts (July 2026)](/blog/openai-gpt-5-6-luna-terra-bedrock-pricing-2026/)
- [Claude Opus 5.5 on AWS Bedrock (September 2026)](/blog/claude-opus-5-5-aws-bedrock-2026/)
- [OpenAI Models, Codex, and Managed Agents on Bedrock](/blog/amazon-bedrock-openai-models-codex-managed-agents/)
- [AWS Bedrock vs OpenAI API for Enterprise](/blog/aws-bedrock-vs-openai-api-enterprise/)
- [AWS Bedrock Cost Optimization: Token Budgets and Model Selection](/blog/aws-bedrock-cost-optimization-token-budgets-model-selection/)

## FAQ

### Are GPT-6 Sol and GPT-6 Luna generally available on Amazon Bedrock?
Yes. AWS announced general availability on September 22, 2026. Sol is the daily model for recurring complex tasks and software development. Luna is the efficient model for high-volume summarization, extraction, classification, and routing. Both support up to 1M tokens of context and explicit prompt caching. GPT-6 Astra, GA September 8, 2026, remains the flagship above them.

### What are the Bedrock model IDs for GPT-6 Sol and Luna?
OpenAI’s Amazon Bedrock guide lists Mantle IDs openai.gpt-6-sol and openai.gpt-6-luna in us-east-1. On bedrock-runtime, use the United States profiles us.openai.gpt-6-sol and us.openai.gpt-6-luna, or the global profiles global.openai.gpt-6-sol and global.openai.gpt-6-luna. Copy the ID from the console if the live card differs.

### How much do GPT-6 Sol and Luna cost on Bedrock?
AWS says both are priced significantly lower than their GPT-5.6 predecessors and does not publish dollar rates in the announcement. OpenAI’s API docs list first-party rates of $2/$10 per million tokens for Sol and $0.10/$0.50 for Luna. OpenAI’s Bedrock guide says commercial Bedrock pricing matches OpenAI direct pricing, and that a region-specific Bedrock service is priced like OpenAI Regional processing. Confirm the Bedrock pricing page before you re-forecast — those first-party numbers are not a Bedrock invoice.

### Should we replace GPT-5.6 Luna and Sol immediately?
No. Route new high-volume lanes to GPT-6 Luna and new recurring complex or coding lanes to GPT-6 Sol only after a frozen-prompt bakeoff. Keep GPT-5.6 Terra and Sol as the prior pins until quality pass rate, latency, and cost per completed task say otherwise. Reserve GPT-6 Astra for the hardest end-to-end jobs.

### Does GPT-6 on Bedrock train on my prompts?
AWS says inference data is not used for model training, and using GPT-6 Sol or Luna does not require opting in to share data with OpenAI. Classifier-flagged abuse-detection traffic is retained by AWS for up to 30 days. Zero data retention is a request through your AWS account team, not the default described for these models.

---

*Source: https://www.factualminds.com/blog/openai-gpt-6-sol-luna-amazon-bedrock-2026/*