v3.10.3

Released October 21, 2026

Makes automatic token increases apply in real time, and restores reliable streaming and prompt caching for Claude models on AWS Bedrock GovCloud.

Confirming this release. When an allocation policy’s automatic increase is enabled, a user’s allowance now rises as soon as usage reaches the configured trigger rather than at the top of the hour; and Claude Code sessions against Claude models on AWS Bedrock GovCloud stream without errors. These changes are largely behind‑the‑scenes — confirm your deployment’s version with your Ask Sage administrator.

⚡ Improvements & fixes

• Automatic token increases now apply in real time. When an organization’s token‑allocation policy is set to automatically increase a user’s allowance at a usage threshold, the increase now takes effect as soon as the user reaches that threshold — automatically, in the background — instead of waiting for the hourly cycle. Because the top‑up can now happen before a user reaches their limit, they’re far less likely to be interrupted.
• Reliable streaming and prompt caching for Claude models on AWS Bedrock GovCloud. Certain requests from Claude Code (and other Anthropic‑compatible clients) to Claude models served on AWS Bedrock GovCloud were being rejected and then silently retried in a way that didn’t reuse cached context — causing errors and higher token usage. Those requests now succeed, stream reliably, and reuse cached context as expected.

🔒 Security & platform

Ask Sage services are delivered as FIPS‑hardened, digest‑pinned, and cosign‑signed images within the FedRAMP Class D authorization boundary, and are continuously rebuilt against the latest patched base images as part of Ask Sage’s continuous‑monitoring commitment.