OpenAI's GPT-5.6 Sol, Terra, and Luna models are now generally available on Amazon Bedrock. The models cover a range of workloads, from autonomous coding to high-volume inference, and are priced at OpenAI's first-party rates.
TL;DR
- OpenAI's GPT-5.6 models are now available on Amazon Bedrock, offering a range of capabilities for different workloads.
- The models are priced at OpenAI's first-party rates and are available in three tiers: Sol, Terra, and Luna.
- Developers can access the models through the OpenAI Responses API on the bedrock-mantle endpoint.
What happened
OpenAI's GPT-5.6 models, including Sol, Terra, and Luna, are now generally available on Amazon Bedrock. These models are designed to handle various workloads, from autonomous coding and long-horizon reasoning to high-volume, latency-sensitive inference.
The models are accessible through the OpenAI Responses API on the bedrock-mantle endpoint. Sol is the flagship reasoning model, Terra balances performance and cost for everyday production work, and Luna is optimized for fast, low-cost inference.
Pricing for the models matches OpenAI's first-party rates, and usage counts toward existing AWS commitments. Developers can select a model, run inferences through the Responses API, and reduce costs with prompt caching.
Why it matters
The availability of GPT-5.6 models on Amazon Bedrock provides developers with a familiar API to access frontier models without operating separate model infrastructure. This can streamline development and deployment processes for AI applications.
The models offer different capabilities and pricing tiers, allowing developers to right-size their workloads for cost and performance. This flexibility can be particularly beneficial for startups and developers working on a variety of projects.
The integration with AWS also provides enhanced security, Regional processing, and cost controls, making it easier for teams to meet data-residency requirements and manage their AI workloads effectively.
Key facts
- GPT-5.6 models are available on Amazon Bedrock, including Sol, Terra, and Luna.
- Sol is the flagship reasoning model, Terra balances performance and cost, and Luna is optimized for fast, low-cost inference.
- Pricing matches OpenAI's first-party rates, and usage counts toward existing AWS commitments.
- Models are accessible through the OpenAI Responses API on the bedrock-mantle endpoint.
- Prompt caching is supported to reduce costs for repeatedly processed context.
- Models support text and image input, text output, and a 272K-token context window.
- Available AWS Regions include US East (N. Virginia), US East (Ohio), and US West (Oregon).
- Developers can control reasoning effort levels, from none to max, to balance performance and cost.
Context
The release of GPT-5.6 models on Amazon Bedrock is part of a broader trend of AI model providers partnering with cloud platforms to offer their models to a wider audience. This collaboration can help accelerate AI adoption and innovation by making powerful models more accessible.
The integration with AWS also highlights the importance of security and data residency in AI deployments. As AI models become more powerful and widely used, ensuring data security and compliance with regional regulations is crucial for businesses and developers.
For developers and startups, the availability of GPT-5.6 models on Amazon Bedrock provides a valuable resource for building and deploying AI applications. The flexibility and cost-efficiency of the models can support a wide range of use cases, from research and development to production workloads.
