GPT-5.5 on AWS Bedrock Ends Azure's Seven-Year OpenAI Exclusivity

The most consequential infrastructure shift of 2026 happened in a 48-hour window in late April. On April 27, Microsoft and OpenAI amended their partnership, ending seven years of Azure exclusivity. On April 28, GPT-5.5, Codex, and Managed Agents powered by OpenAI went live on Amazon Bedrock in limited preview. GPT-5.5 matches GPT-5.4 per-token latency while reaching 93.6% on GPQA Diamond and 82.7% on Terminal-Bench 2.0. API pricing sits at $5 per million input tokens and $30 per million output tokens, with the Pro variant at $30/$180 for highest-end reasoning tasks. More than 4 million people now use Codex weekly. The strategic significance goes beyond the model itself. AWS customers can now invoke GPT-5.5 through the same Bedrock APIs already used for Anthropic, Meta, and Amazon's Nova family — and apply usage toward existing AWS spending commitments. Azure remains the primary OpenAI partner through 2032, but the lock-in is gone. Enterprises with pre-negotiated AWS spend commitments can route OpenAI workloads through a single cloud account without establishing a separate Azure relationship. Combined with DeepSeek V4-Flash pricing at $0.14 per million input tokens, the AI inference market has entered a genuine multi-provider commodity phase. For engineering teams, this means architecting for model portability is no longer optional — it is the rational economic choice. The days of single-vendor AI dependency are effectively over.