OpenCode v1.18.33 (2026-09-28) makes Cloudflare AI Gateway models honor provider response/stream timeouts, reports MCP browser launcher exits, redacts credentials in debug config output, and aligns Gemini thinking defaults/effort options with supported controls across model generations.

Key Takeaways

  • ✓Shipped v1.18.33 — CF AI Gateway timeouts honored
  • ✓MCP browser launcher immediate-exit failures reported
  • ✓Debug config redacts credentials/sensitive headers
  • ✓Gemini thinking defaults/effort aligned across generations
🧭

Finished reading? Explore benchmark rankings & pricing

Real-world SWE-bench scores & $20/mo vs API cost break-even calculator

🏆 Leaderboard→💰 Compare Plans↗
🔗

Project Links & Resources

Direct Access
Direct access to official project resources and documentation
🐙
GitHub Release v1.18.33
github.com/anomalyco/opencode/releases/tag/v1.18.33
↗
🐙
anomalyco/opencode repository
github.com/anomalyco/opencode
↗
📖
OpenCode documentation
opencode.ai/docs
↗
📖
OpenCode homepage
opencode.ai
↗
🐙
PR #51549 CF AI Gateway timeouts
github.com/anomalyco/opencode/pull/51549
↗
🐙
#50841
github.com/anomalyco/opencode/pull/50841
↗
🔬

In-Depth Technical Analysis

Core Background & Industry Pain Points

OpenCode behind Cloudflare AI Gateway or Gemini thinking often fails silently: ignored timeouts hang streams, MCP browser launch failures stay quiet, and debug logs leak API keys.

Architecture Highlights & Internals

v1.18.33 applies provider response/stream timeouts on CF AI Gateway paths (#51549), surfaces immediate MCP browser launcher exits, redacts credentials in debug config, and aligns Gemini thinking defaults/effort with supported controls (#50841).

Authoritative Benchmarks & Measured Scores

No public SWE-bench/latency table—validate timeout abort, MCP launch errors, and credential redaction on your stack.

Developer Hands-on Guide

Upgrade via docs to v1.18.33. See the release notes.