New features, improvements, and fixes. We ship fast and often.
Searchable Guides Library
AddedThe Guides page is now a searchable, categorized resource library — find setup and how-to articles fast.
Global Financial Data
AddedNew data integrations for SEC filings, the World Bank, and the IMF.
GPT-5.6 Arrives
AddedGPT-5.6 (Sol, Terra, and Luna) is now available in the ChatGPT-subscription model picker.
WhatsApp on Hermes
FixedWhatsApp now connects reliably on Hermes bots.
Share a Live Browser Link
AddedShare a one-tap Live Browser link so your bot can log in for you — no passwords in chat.
AddedDeepSeek and OpenAI now work as direct bring-your-own-key models.
Live Browser Logins & Platform Guides
AddedYour bot hands you a Live Browser button for account logins instead of asking for your password.
AddedNew guides: app builder, static site hosting, custom domains, cloud terminal & files, cron, and managed models.
AddedBring your own OpenAI key for chat on OpenClaw.
AddedConnect any custom OpenAI-compatible API provider from the API Keys hub.
Telegram Groups & More Trial Models
AddedTelegram group mention-only mode — the bot replies only when you @mention it.
ImprovedMore trial models to choose from, including GLM-5.2 and Kimi.
AddedMore ready-to-use skills added to the marketplace.
OpenClaw on Your Phone
AddedNative iPhone and Android apps are here — run and chat with your AI agent right from your phone.
AddedMore ready-to-use skills added to the marketplace.
Smoother Big Downloads
ImprovedLarge file downloads now resume instead of stalling near the end.
Portfolio Tracking & Skills in Chinese
AddedNew read-only IBKR skill lets your bot check your holdings and cash balances.
ImprovedEvery marketplace skill now has a Chinese description.
Reliable File Delivery
FixedYour bot now reliably delivers files in web chat, even on fast streaming models.
Bring Your Own OpenAI Key
AddedUse your own OpenAI key for chat on Hermes and pick your primary chat model.
Faster Claude & Clearer Messages
ImprovedClaude with your own Anthropic key now connects directly to Anthropic for faster, more reliable replies on Hermes.
ImprovedClearer "usage limit reached" message when a model hits its cap.
Faster Memory + Sharper Images
ImprovedBot memory search is now several times faster — quicker recall with the same quality
ImprovedFree-trial image generation upgraded to a newer, higher-quality model
ImprovedCleaner Hermes chat commands — /help, /new and /platforms are compact and reliable
Coding Workspace + App Building
AddedNew Coding Workspace — a disposable cloud dev box with AI coding agents (Claude, Codex, OpenCode, Aider) built in, all driven from your browser terminal. Free-trial users can try it too
AddedYour bot can now build and run Android, iOS, and Expo apps — the build toolchain installs automatically
AddedNew Apps page in your dashboard shows everything your bot can create
Bigger Uploads, Whole Folders
ImprovedUpload multiple files — or an entire folder — at once in the Files manager
ImprovedFile upload limit raised from 5 MB to 50 MB
Your Bot Can Send Big Files
AddedWhen a file is too large for the chat app, your bot sends it as a secure, time-limited download link
AddedFiles your bot creates now appear as clickable download links right in web chat
Inline Media + Bigger Backups
ImprovedImages and SVGs your bot makes now render inline in web chat
ImprovedKeep typing in web chat while the AI is still replying
ImprovedExport Backup now supports larger archives (up to 2 GB) with a one-click Select all
Nous Portal + Full MiniMax Catalog
AddedBring your own Nous Portal key — works on both OpenClaw and Hermes, with a searchable live model picker
AddedFull MiniMax direct catalog — all 8 models now available in the picker
ImprovedImage model picker redesigned into a clean two-pane provider→model layout, matching the text picker
Auto-Wake Your Bot + Gemini 2.5 Flash
AddedSend a web-chat message to a sleeping bot and it wakes itself up automatically — no extra clicks
AddedGemini 2.5 Flash and Flash-Lite added to the model menu
ImprovedBrowse and edit your bot's Website page anytime, even before an instance is running
Visual File Browser
AddedBrowse, view, and manage your bot's files in a clean two-pane visual browser — open to all paid plans
ImprovedHermes instance card now opens straight into the file browser
Redesigned Model Picker
AddedNew two-pane model picker: every saved BYOK key gets its own pane with auth badges, plus always-on search
ImprovedImage, video, and music pickers now show the live, daily-refreshed FAL catalog
More BYOK Options
AddedBring your own OpenCode Go key with a built-in models.dev model picker
AddedMiniMax Coding Plan keys now supported
AddedAdd Tavily as a per-instance web-search provider with your own key
Hermes WhatsApp + Image Support
AddedConnect WhatsApp to your Hermes bot
AddedHermes chat now supports file upload and image paste, and images work with any primary model
AddedSlash-command menu in Hermes chat for quick actions
ImprovedRedesigned deploy flow — a clear 5-step checklist replaces the single spinner
Hermes Browser Tool + Trending FAL Models
AddedHermes Agent can now browse the web with a built-in Chrome — same browser tool OpenClaw has
AddedTrending FAL music and video models in the picker, refreshed daily with live pricing
AddedAnnual upgrade button right inside the billing card — switch from monthly without leaving the page
ImprovedMore reliable WhatsApp connect — handles cases where the gateway stays silent after pairing
Live OpenRouter Top-20 + 9 New BYOK Tiles
AddedModel picker shows live OpenRouter top-20, refreshed hourly so trending models surface fast
AddedBring your own keys for DeepSeek, Moonshot, Groq, Qwen, Z.AI, Runway, BytePlus, ComfyUI, and Replicate
AddedGPT-5.5 available in the ChatGPT Subscription model picker
ImprovedModel dropdown collapses less-popular models per provider so the list stays short
Hermes × Composio + Native WeChat QR Login
AddedHermes routes Google Workspace, Notion, and Slack through Composio for richer skills
AddedHermes WeChat now uses the native iLink Bot QR login — scan once and you're paired
AddedOpenAI Codex section visible in the Hermes model dropdown
ImprovedGoogle Drive uploads now support 5–50 MB files via resumable upload
BYOK Primary Picker + Faster Dashboard
AddedPick your primary model directly on the API Keys page — your BYOK choice routes everything by default
ImprovedDashboard renders your instance list on first paint — no round-trip flash
ImprovedDeploy card defaults to the Hermes tab when you already have an OpenClaw instance
Gemma on Google BYOK
AddedGoogle Gemma models available with your Google BYOK key
Hermes Post-Deploy Polish + Web Terminal Everywhere
AddedWeb terminal now works on every container node — faster and more reliable
AddedHermes shows the same post-deploy modal as OpenClaw, with a Connect Platforms shortcut
AddedHermes ChatGPT Subscription setup guide added to the help center
ImprovedWelcome modal opens chat inline instead of jumping to a new tab; WeChat connect shows a 2–4 min ETA with step hints
OpenClaw 2026.5.7 + WeChat 2.4.3
ImprovedBumped OpenClaw runtime to 2026.5.7 and WeChat plugin to 2.4.3
AddedCollapsible sidebar in the Hermes dashboard for more chat space
Annual Plans + Hermes Parity
AddedAnnual plans — $60/yr Lite, $200/yr Pro. Switch from monthly any time with prorated refund
AddedSkills marketplace and Edit + Revert UI now work on Hermes too — same one-click rollback OpenClaw has
Live Tool Progress + BYOK to Hermes
ImprovedWeb chat shows live tool progress (e.g. "searching the web") instead of a static "Thinking…"
AddedMiniMax, FAL, Google, and Anthropic BYOK keys propagate to your Hermes containers
AddedBots can now send generated images, video, music, and docs straight into the chat
Hermes on Feishu + Image Generation
AddedConnect Hermes Agent to Feishu and Lark
AddedImage generation for Hermes — pick from OpenRouter, Google, and MiniMax models
xAI Grok BYOK
AddedBring your own xAI (Grok) API key — Grok 4.20 and Grok 4.1 Fast in the model picker
Smarter Web Terminal
AddedCopy / Paste / Clear toolbar, Copy All button, and live ping in the terminal status bar
AddedOpt-in low-latency mode — predictive local echo for laggy SSH connections