v3.9.7

Released October 1, 2026

Adds two new models — Claude Sonnet 5.5 and GPT‑6.1 Sol.

Upgrade note. This release corrects how the Anthropic‑compatible API applies the reasoning‑effort setting. Rolling back to the previous version would reintroduce the earlier behavior (effort ignored), so a rollback is not recommended — but it is otherwise safe (no data or schema changes).

Confirming this release. Claude Sonnet 5.5 and GPT‑6.1 Sol appear in the model picker. New models are visible only when enabled for your deployment, within your compliance boundary, and for your organization — if you don’t see one, contact your Ask Sage administrator.

🚀 New models

• Claude Sonnet 5.5 — government & DoD. Anthropic’s Claude Sonnet 5.5 is now available on government and DoD (IL4 / IL5) deployments via AWS Bedrock GovCloud, with a 1M‑token context window, and is flagged for CUI / ITAR / EAR data controls.
• GPT‑6.1 Sol — commercial. OpenAI’s GPT‑6.1 Sol is now available on commercial deployments (Azure AI Foundry), with a ~1M‑token context window, vision input, tool use, reasoning, and native web search.

⚡ Improvements & fixes

• Reasoning effort now takes effect on the Anthropic‑compatible API. The reasoning‑effort setting was previously dropped before reaching the model; it is now passed through, so requests honor the effort you specify.

🔒 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.