Released: April 24, 2026
Generated: April 25, 2026
Sources: GitHub Release Notes, OpenClaw Documentation
OpenClaw v2026.4.23 is a feature-heavy release delivering native image generation through multiple providers, advanced agent orchestration with forked context inheritance, and deeper Codex OAuth integration.
openai/gpt-image-2 works without OPENAI_API_KEYOPENROUTER_API_KEYImpact: Removes friction for users already authenticated via Codex OAuth or OpenRouter — no additional API key management needed.
sessions_spawn runs can now inherit the requester transcriptUse Case: Complex multi-step workflows where a child agent needs full conversation history to make contextually aware decisions, while simple tasks stay isolated.
timeoutMs for image, video, music, and TTS generation toolsmemorySearch.local.contextSize with 4096 defaultrequest_user_input prompts route back to originating chat/status stays simple while gateway logs explain auto-selection and Pi fallback reasons0.70.0gpt-5.5 catalog metadata for OpenAI and OpenAI Codexgpt-5.5-pro forward-compat handling retainedNote: This aligns OpenClaw with OpenAI's latest model catalog on the same day GPT-5.5 shipped — maintaining the tight integration pattern seen in previous releases.
# Update OpenClaw
npm update -g openclaw
# Or via your package manager
openclaw --version # Should show 2026.4.23
sessions_spawn parametersStrategic Observation: OpenClaw is iterating faster than most commercial platforms, shipping meaningful features daily. The Codex OAuth integration suggests a bet on OpenAI's enterprise distribution strategy.