OpenClaw's latest release (published April 6) shipped a few changes that matter immediately for people running personal or self-hosted agent stacks.
What changed (high-impact items)
1) Native media generation tools are now first-class
The release add...
Anthropic has reportedly changed how Claude Code subscription usage works for third-party harnesses, and OpenClaw is directly in the blast radius.
According to widely shared reports and user-posted notices, Claude subscription limits no longer apply...
OpenClaw 2026.3.28 introduced a set of changes that matter for people running self-hosted/personal agents in production.
What changed (and why it matters)
1) Approval gates can now happen before a tool call runs
The release adds async approval hooks...
Based on numerous requests, we’ve added a new feature to Clawly: you can now jump directly into the terminal where your OpenClaw agent is running and use the openclaw CLI yourself.
This gives advanced users more direct control for debugging, configur...
OpenClaw's latest release added a practical compatibility layer many self-hosters have been waiting for: OpenAI-style endpoints directly on the Gateway.
In the release notes, OpenClaw explicitly calls out support for:
GET /v1/models
POST /v1/embeddi...
OpenClaw has a fresh March 2026 update cycle worth acting on now if you self-host. The latest release line (v2026.3.23 on GitHub, 2026.3.23-2 on npm) ships several reliability and security-relevant fixes, especially around plugins, auth flows, and ch...
OpenClaw shipped a recovery release as v2026.3.13-1 on GitHub after a broken v2026.3.13 release path. The key detail: npm stays on 2026.3.13 (no -1 suffix there), while GitHub uses -1 because immutable releases cannot be overwritten.
Why this matters...
OpenClaw 2026.3.8 is out, and this release is a practical upgrade for people running agents in production.
What changed
New backup commands: openclaw backup create and openclaw backup verify
Stronger external/untrusted content handling
Reliability f...
We've added two new AI models from xAI — Grok 4 and Grok 4.1 Fast — available on both BYOK and Managed plans starting today.
The models
Grok 4 is xAI's flagship reasoning model. It offers strong analytical capabilities, a 256K token context window, a...
Anthropic just released Claude Sonnet 4.6, the successor to Sonnet 4.5 — and starting today, all new agent deployments on Clawly will use it by default.
What's new
Sonnet 4.6 is a drop-in upgrade over Sonnet 4.5. It delivers stronger reasoning, bette...