OpenAI GPT-5.5, GPT-5.4 and Codex Now Generally Available on Bedrock

OpenAI GPT-5.5, GPT-5.4 models, and Codex are now generally available on Amazon Bedrock. The availability follows a preview in What’s Next with AWS 2026 and provides access to frontier models plus a coding agent for software development.

GPT-5.5 and GPT-5.4 deliver capabilities for demanding workloads. According to OpenAI, the models are excellent for coding, reasoning, agentic workflows, and complex professional work. GPT-5.5 targets the hardest customer workloads while GPT-5.4 delivers the best price-performance.

Both models are called through the Responses API on Amazon Bedrock’s next-generation inference engine. That engine is built for high performance, reliability, and security.

Codex powers AI-assisted development at scale. Codex is the OpenAI coding agent for AI-powered software development. According to OpenAI, more than 4 million developers use Codex every week to write, refactor, debug, test, and validate code across large codebases.

With GPT-5.5 powering inference, Codex introduces a new class of intelligence optimized for complex, long-horizon developer workflows. It works through the Codex App, the Codex CLI, and IDE integrations with Visual Studio Code, JetBrains, and Xcode. All model inference routes through the Responses API on Amazon Bedrock.

Bedrock meets enterprise data and cost requirements. For customers with data residency requirements, all processing stays within the Bedrock Region selected. Customers pay per token with no seat licenses and no per-developer commitments.

Programmatic access uses the OpenAI Responses API. Developers access the models programmatically by calling bedrock-mantle endpoints through the OpenAI SDK or command-line tools such as curl. Sample Python code installs the OpenAI SDK with pip, sets environment variables for OPENAI_BASE_URL, OPENAI_API_KEY and BEDROCK_OPENAI_MODEL_ID, then creates a client to call responses.create with developer and user roles.

One provided example prompts a software engineer with AWS cloud knowledge to design a distributed architecture in Python supporting 100k requests per second across multiple geographic regions. The Responses API suits model-managed multi-turn state, hosted tools, function tools, richer tool orchestration, and background or long-running work.

Codex tools support flexible Bedrock authentication. Users download the Codex CLI, Codex App or Codex VS Code extension to begin. Codex supports two Bedrock authentication pathways: Amazon Bedrock API key or AWS SDK credential chain. It reads AWS_BEARER_TOKEN_BEDROCK first if set and otherwise falls back to the AWS SDK credential chain.