AWS Adds UltraFast Mode for OpenAI GPT-6 Astra on Amazon Bedrock
AWS has introduced UltraFast mode for OpenAI GPT-6 Astra on Amazon Bedrock. This premium speed tier delivers up to 6x faster inference, reaching 300 tokens per second, designed for latency-sensitive AI applications. It benefits developers building real-time coding assistants, interactive agents, and customer-facing experiences requiring swift, high-quality responses. Users can access this feature through the Amazon Bedrock console or APIs, leveraging AWS controls for security and governance.
Features (1) ›
- UltraFast Mode for OpenAI GPT-6 Astra on Amazon Bedrock
AWS has made UltraFast mode available for OpenAI GPT-6 Astra on Amazon Bedrock. This premium speed tier is optimized for workloads where inference speed is critical, providing up to 6x faster inference and processing up to 300 tokens per second.
https://aws.amazon.com/about-aws/whats-new/2026/09/openai-gpt-6-astra-ultrafast-on-amazon-bedrock/
Related releases
- Amazon Bedrock Adds Z.ai's GLM 5.3 Model for Agentic Software Engineering AWS What's New ·
- Amazon Nova 2.5 Sonic GA: Enhanced Reasoning for Real-time Voice Agents AWS What's New ·
- AWS Weekly Roundup details Bedrock Agents, Q3 Lifecycle Updates, and AI Models AWS News Blog ·
- Amazon Bedrock AgentCore Gateway supports private TLS certificates for VPC endpoints AWS What's New ·
- Amazon Bedrock adds Claude Opus 5.5 and Sonnet 5 in UK (London) region AWS What's New ·
- AWS Introduces AI-Driven EKS Migration Assessment Solution Using Bedrock AWS Containers Blog ·