OpenAI's GPT-5.5 and Codex Reach AWS Bedrock: India's Read

The News

OpenAI's most capable systems have arrived inside Amazon's cloud. On 1 June 2026, AWS confirmed that GPT-5.5, GPT-5.4 and the company's Codex coding tool are now generally available on Amazon Bedrock, the managed service through which enterprises already run their machine-learning workloads.

The pitch is procurement, not novelty. Companies can now reach OpenAI's frontier models using the security controls, governance frameworks and purchasing agreements they have already signed with AWS, rather than wiring up a separate vendor relationship. According to OpenAI, the goal is to help customers move faster from evaluation to production.

Two numbers anchor the announcement. First, pricing: AWS states that the rates "match OpenAI first-party rates, and usage counts toward existing AWS commitments", meaning spend on these models can be drawn against committed cloud budgets. Second, capability: GPT-5.5 is positioned for "agentic coding, data analysis, and multi-step autonomous tasks" and runs on Bedrock's next-generation inference engine. Codex itself is reachable through the Codex App, the Codex CLI and IDE plug-ins for Visual Studio Code, JetBrains and Xcode.

Why It Matters

This is a distribution story dressed as a product launch. OpenAI built its enterprise base largely through its own API and a deep tie to Microsoft Azure. Putting the same frontier models on a direct competitor's platform widens the funnel considerably, and it signals that model makers increasingly see neutrality across clouds as a commercial necessity rather than a betrayal of an anchor investor.

The last comparable moment was Anthropic's expansion onto Bedrock in 2023, which turned a research-heavy lab into a default enterprise option almost overnight by meeting buyers where their data already lived. The lesson held: in enterprise software, the winning model is often the one that is easiest to procure, not the one that tops a benchmark by two points. By matching first-party pricing, OpenAI removes the usual reseller penalty and bets that volume through AWS outweighs the margin it might have captured selling direct.

There is a defensive angle too. With multiple frontier labs now sitting side by side inside Bedrock, the cloud provider becomes the layer that captures loyalty, while individual models grow more swappable. That commoditisation pressure is exactly what every foundation-model company is racing to outrun.

Indian Angle

For Indian enterprises, the meaningful detail is where Bedrock runs. AWS operates regions in Mumbai and Hyderabad, so a bank, insurer or fintech bound by RBI data-localisation rules and the Digital Personal Data Protection Act can, in principle, consume frontier OpenAI models without routing sensitive data outside India. That removes one of the largest blockers to generative-AI adoption inside regulated BFSI shops, where compliance teams have repeatedly vetoed direct foreign API calls.

The pricing structure also lands differently in rupee terms. Many large Indian buyers, from Tata and Infosys to digital-first lenders, already sit on multi-year AWS commitments. Letting GPT-5.5 usage count against those commitments converts a fresh budget request into an existing line item, which is often the difference between a pilot and a rollout.

It sharpens the question for India's own model builders. Sarvam AI and Ola's Krutrim have pitched sovereign, India-first models partly on the argument that global frontier systems are hard to deploy compliantly at home. Easy, in-region access to GPT-5.5 narrows that gap and raises the bar: domestic labs will now need to compete on cost, Indic-language depth and genuine data sovereignty rather than on availability alone.

FAQ

When did this take effect?

General availability began on 1 June 2026, when AWS announced that GPT-5.5, GPT-5.4 and Codex were live on Amazon Bedrock for all eligible customers.

Does using these models cost more on AWS?

No. AWS says pricing matches OpenAI's first-party rates, and the usage counts toward a customer's existing AWS spending commitments rather than being billed as a separate contract.

What does this mean for Indian regulated firms?

With Bedrock available in the Mumbai and Hyderabad regions, banks and insurers can potentially use frontier OpenAI models while keeping data in India, easing RBI localisation and DPDP Act concerns that previously blocked adoption.

How can developers access Codex?

Codex is available through the Codex App, the Codex CLI and IDE integrations for Visual Studio Code, JetBrains and Xcode, alongside the GPT-5.5 and GPT-5.4 models on Bedrock.

This story was reported by OpenAI. Read the full original coverage at OpenAI.