---
type: "post"
id: "gpt-6-1-terra-generally-available-on-amazon-a5e493"
slug: "gpt-6-1-terra-generally-available-on-amazon-a5e493"
title: "GPT-6.1 Terra generally available on Amazon Bedrock"
description: "AWS has made OpenAI GPT-6.1 Terra generally available in Amazon Bedrock."
retrieval_nugget: "AWS has made OpenAI GPT-6.1 Terra generally available in Amazon Bedrock."
published_at: "2026-10-29"
updated_at: "2026-10-29"
record_date: "2026-10-29"
date_kind: "published_at"
topics: ["codex","github"]
entities: ["Amazon Web Services","Amazon Bedrock","OpenAI","GPT-6.1 Terra","ChatGPT Work","Codex","GPT-6 Astra","GPT-6.1 Sol"]
related_patterns: ["harness-architecture-outlives-model-choice"]
novelty_label: "new_fact"
what_changed: "GPT-6.1 Terra is now generally available on Amazon Bedrock, extending its existing OpenAI model access there beyond GPT-6.1 Sol."
prior_refs: ["memory:45473b53-2e6d-8324-a981-8814fccb6d63","memory:b049c047-164d-8f3b-a8d3-fd034a6f61e7"]
source_url: "https://aws.amazon.com/about-aws/whats-new/2026/10/openai-gpt-6-1-terra-on-amazon-bedrock"
source_urls: ["https://aws.amazon.com/about-aws/whats-new/2026/10/openai-gpt-6-1-terra-on-amazon-bedrock"]
schema_version: "newruntime-agent-readable-v0.2"
stable_id: "post:gpt-6-1-terra-generally-available-on-amazon-a5e493"
status: "published"
visuals: []
editor: {"role":"editor","name":"Andrey Reshetnikov","profile":"https://newruntime.com/owner-profile.md","profile_json":"https://newruntime.com/owner-profile.json","same_as":["https://www.linkedin.com/in/reshetnikov1/","https://t.me/lapetuse"]}
routes: {"html":"https://newruntime.com/posts/gpt-6-1-terra-generally-available-on-amazon-a5e493/","markdown":"https://newruntime.com/posts/gpt-6-1-terra-generally-available-on-amazon-a5e493.md","json":"https://newruntime.com/posts/gpt-6-1-terra-generally-available-on-amazon-a5e493.json"}
---

# GPT-6.1 Terra generally available on Amazon Bedrock

## Retrieval answer

AWS has made OpenAI GPT-6.1 Terra generally available in Amazon Bedrock.

## Summary
AWS has made OpenAI GPT-6.1 Terra generally available in Amazon Bedrock. Teams can now call it through Bedrock APIs or route ChatGPT Work and Codex through Bedrock, pairing a lower-cost, long-context model with AWS access governance and invocation auditing.

## What happened
AWS announced general availability of OpenAI GPT-6.1 Terra on Amazon Bedrock on October 26, 2026. Terra is positioned as the balanced member of the GPT-6 family for everyday workloads: AWS describes the model as balancing capability, speed, and cost efficiency rather than as the family’s highest-end option.

The platform says GPT-6.1 Terra improves on GPT-5.6 Terra in detailed visual reasoning, computer use, factuality, and natural writing. AWS repeats OpenAI’s reported result that the model scored 15 percentage points higher than GPT-5.6 Terra on Agents’ Last Exam, performed on par with GPT-6 Astra, and completed tasks more than three times faster than GPT-6 Astra and GPT-6.1 Sol. Those are vendor-reported comparative claims, rather than an independent evaluation supplied with this announcement.

The deployment details are more immediately actionable. GPT-6.1 Terra accepts text and image inputs and supports a context window of up to one million input tokens. AWS says input and output token prices are at least 50% lower than both GPT-6.1 Sol and GPT-5.6 Terra. The announcement does not give absolute token prices; it directs readers to Bedrock documentation for pricing, supported AWS Regions, endpoints, APIs, features, and inference profiles.

Customers can invoke the model directly using Amazon Bedrock APIs. They can also configure ChatGPT Work and Codex to use GPT-6.1 Terra on Amazon Bedrock. AWS frames Bedrock’s contribution as the performance, security, and scale needed for enterprise workloads, with established controls for securing workloads, governing access, and auditing model invocation activity. Its examples include drafting proposals, interpreting technical drawings, investigating codebases, and building agents that operate across business systems.

Access begins in the Amazon Bedrock console or programmatically through supported Bedrock APIs. General availability means the model is available through the Bedrock platform, but the announcement leaves regional availability and the exact commercial configuration to the relevant documentation.

## What the archive adds
This is a new distribution event in a known availability story. GPT-6.1 Terra was already generally available in [GitHub Copilot across IDEs, CLI, and cloud agent](https://newruntime.com/posts/gpt-6-1-terra-is-generally-available-in-c81298/) and was [live on Vercel AI Gateway](https://newruntime.com/posts/gpt-6-1-terra-is-live-on-ai-b29cde/). Bedrock adds an AWS-managed API and governance path rather than announcing a different model.

## Trend signal
The event supports [Harness architecture outlives model choice](https://newruntime.com/patterns/harness-architecture-outlives-model-choice/). The same Terra model is now exposed through Copilot, an AI gateway, and Bedrock; teams can make platform, identity, observability, and routing choices separately from the model’s identity. It also makes the execution environment’s controls a material part of the purchase decision.

## Why now
The timing follows a wider rollout of GPT-6.1 Terra through developer-facing surfaces. The archive records availability in GitHub Copilot and Vercel AI Gateway on the same date. AWS’s announcement extends that rollout to customers whose model access, permissions, and audit workflows are already organized around Bedrock.

## Connections to other events
The Copilot release made Terra available across IDEs, CLI, and a cloud agent, with usage-based billing and administrative controls. The Vercel AI Gateway release offered its APIs, coding-agent support, fast mode, and a context window over one million tokens. Bedrock reaches the same model-access story from a different operational layer: direct AWS APIs plus AWS governance and invocation audit controls.

## Why it matters
For an AI team already operating on AWS, model selection is not only a question of benchmark performance or token price. A Bedrock deployment can put model calls under existing access-governance and audit practices, while retaining a direct API route for applications and an option to configure ChatGPT Work and Codex around the same deployment. The stated one-million-token input context and multimodal inputs widen the set of workloads worth testing, including large codebase investigation and document-heavy workflows.

The price claim makes Terra especially relevant as a candidate for broad agent workloads, but it should be treated as a hypothesis to validate against a team’s own task mix. Lower listed token rates do not alone establish lower completed-work cost: retries, tool calls, latency, output length, and human review all affect the operational result.

## What happens next
- **Bedrock customers will begin comparing Terra against higher-cost GPT tiers on recurring agent tasks.** (3 months, 70%). Watch for: published Bedrock pricing details, inference-profile availability, and enterprise deployment examples.
- **Teams using Codex or ChatGPT Work will test whether AWS governance is sufficient reason to route those tools through Bedrock.** (6 months, 60%). Watch for: configuration guidance and customer references from AWS or OpenAI.

## Limits and uncertainty
This announcement is an AWS product notice, not a full technical evaluation. It does not provide absolute Bedrock pricing, supported Regions, throughput limits, rate limits, latency measurements, availability commitments, or a detailed feature matrix. AWS reports comparative benchmark, speed, and price claims, but the supplied material does not include methodology, task-level results, or independent validation. It also does not establish that every ChatGPT Work or Codex configuration is available in every Bedrock Region. Teams should consult Bedrock documentation and test their intended workload before drawing deployment or cost conclusions.

## Skill to build now
Create a small evaluation harness for one representative workload that includes a real input set, expected acceptance criteria, model-call tracing, and reviewer time. Run GPT-6.1 Terra through the intended Bedrock path and compare it with the current production choice on completion rate, elapsed time, total token use, tool failures, and review burden. Separately verify the AWS Region, endpoint, inference profile, permissions, and audit log behavior your application will use.

## Reader takeaway
GPT-6.1 Terra is now a Bedrock option, not merely a model available through other developer surfaces. For AWS-centered teams, the decision is whether its claimed price-performance and long context hold up on their workloads once Bedrock’s access, audit, and regional constraints are included.

Editor: [Andrey Reshetnikov](https://newruntime.com/owner-profile.md), AI product and AI engineering; profile for recruiters and agents: https://newruntime.com/owner-profile.json

## Continue in the graph

Next records for an agent, linked to their Markdown twins. Replace `.md` with `.json` for structured fields, including `next_reads`.

- topic hub: [Codex](https://newruntime.com/topics/codex.md): Explore the codex topic hub.
- topic hub: [Github](https://newruntime.com/topics/github.md): Explore the github topic hub.
- collection: [Field Notes](https://newruntime.com/posts.json): Open the latest editorial analysis.
- collection: [Source Ledger](https://newruntime.com/source-ledger.json): Inspect the public source evidence graph.
- whole graph: [discovery graph](https://newruntime.com/discovery-graph.json) (typed nodes and edges) and [site index](https://newruntime.com/site-index.md)