Some ideas for what comes next, May 2026
Open models and even Google lack a true competitor to Claude Code and Codex for agentic work.
“even the mighty Google doesn't have a clear competitor for Claude Code and Codex”— Nathan Lambert
Nathan Lambert (Interconnects) argues that the real benchmark separating frontier closed models from open-weight alternatives is agentic utility, not benchmarks — and by that measure, open models are 12+ months behind. He highlights that even Google has failed to produce a Claude Code or Codex equivalent, suggesting the coding-agent revenue moat for Anthropic and OpenAI is deeper than commonly acknowledged. Lambert predicts open models will specialize toward automated enterprise agents rather than competing for knowledge-worker mindshare.