← Back
AWS ML Blog

Bring near-Astra intelligence to everyday work with GPT-6.1 Sol on Amazon Bedrock

•5 min read•
#amazon#bedrock#llm#agents
Level:Advanced
For:AI Engineers
✦TL;DR

GPT‑6.1 Sol is now generally available on Amazon Bedrock, delivering near‑Astra reasoning for coding, computer use, and professional workflows. It matches GPT‑6 Astra on the DeepSWE v1.1 benchmark at roughly one‑fifth the cost per task and beats the prior GPT‑6 Sol best score by 6.4 percentage points while using less reasoning effort. Engineers can tap the model via Codex, the Agent Toolkit for AWS, and Bedrock APIs to build agentic tools that synthesize documents, coordinate services, and automate development cycles.

⚡ Key Takeaways

  • Cost efficiency: roughly one‑fifth the cost per task versus GPT‑6 Astra on DeepSWE v1.1.
  • Performance gain: +6.4 percentage points over the best GPT‑6 Sol result with lower reasoning effort.
  • Integration point: Codex can be configured to use GPT‑6.1 Sol on Bedrock for end‑to‑end software‑engineering tasks.
  • Tooling support: Agent Toolkit for AWS links Codex to AWS docs, APIs, and services via a single terminal command.
  • Limitation handling: Model now better recognizes tool failures, restrictions, or missing information to allow early human intervention.
💡 Why It Matters

The reduced per‑task cost and higher reasoning quality let production teams scale agentic automation without exploding inference spend, while built‑in failure awareness improves safety and reliability of tool‑driven workflows.

✅ Practical Steps

  1. Configure Codex to target GPT‑6.1 Sol on Amazon Bedrock for investigation, implementation, and testing phases.
  2. Use the Agent Toolkit for AWS to connect Codex to AWS documentation, APIs, and services with a single terminal command.
  3. Invoke supported Amazon Bedrock APIs to build internal tools that synthesize documents or coordinate multi‑system agent workflows.

Want the full story? Read the original article.

Read on AWS ML Blog ↗

More like this

Censored vs. Truncated Regression in Python, Explained

Towards Data Science•#llm

New in AWS Quick: More capable flows, an updated web experience, and agents that work together

AWS ML Blog•#agents

How Google Duplex evolved to also call businesses for you in Gemini

Google Research•#agents

Q&A: Enabling more human-like interactions and understanding in multimodal foundation models

MIT News AI•#llm

EXPLORE AI NEWS

Daily hand-picked stories on LLMs, RAG, agents and production AI — curated for engineers who ship.

BROWSE NEWS

GET THE WEEKLY DIGEST

Join engineers getting the Monday signal-over-noise AI breakdown. No spam, unsubscribe anytime.

LEARN AI ENGINEERING

Curated courses, research papers, repos and tutorials built for engineers leveling up in AI.

START LEARNING