Ship agents faster with expanded model choice, voice agents, and continuous optimization
Microsoft Foundry now lets teams pick from new GPT‑6 and Claude models, build voice agents, and auto‑optimize them in production.
Why Now
Microsoft announced new model support (GPT‑6, Claude Opus 5.5) and voice agent preview in Foundry Agent Service, expanding its AI agent platform.
What Happened
Foundry now includes the full GPT‑6 family and Claude Opus 5.5, offering more model choices. Voice agents are in public preview, supporting 80+ languages and 140+ locales. Tool search reduces token usage by >60% for 50‑tool sets and >97% for 1,000‑tool sets. Long‑running resilience and Agent Framework add checkpointing and background workflow support.
Why It Matters
Teams can switch models without rebuilding architecture, speeding innovation and reducing cost. Voice integration lowers friction for conversational AI, while tool search cuts latency and token costs. Resilience features enable more complex, real‑world workflows in production.
The Limitation
The voice preview and new models are still in public preview; performance and stability may vary, and full pricing details are not yet released.
What You Can Do
Try the Foundry Agent Service voice preview and benchmark your own workload against GPT‑6 or Claude Opus 5.5.
Source
Read original sourceWhy we picked this
Discusses AI agents, model choice, voice agents, and optimization – core AI content.