LLM Hacker News May 4, 2026 1 min read
DeepClaude keeps Claude Code's complete agent loop — file editing, bash, subagent spawning — while routing API calls to DeepSeek V4 Pro or other backends, cutting output token costs from $15/M to $0.87/M.
#cost-optimization
RSS FeedLLM X/Twitter Apr 11, 2026 2 min read
Claude said on April 9, 2026 that the advisor strategy is now in beta on Claude Platform. The new tool lets Sonnet or Haiku call Opus for planning help inside a single Messages API request, which Anthropic says raised SWE-bench Multilingual by 2.7 points while cutting cost per task by 11.9% versus Sonnet alone.