📰 Clawdify News & Updates

Stay up-to-date with the latest features, integrations, and OpenClaw engine upgrades automatically applied to your Clawdify agents.

🧠 OpenClaw v2026.4.11 — Already Live on Clawdify

April 12, 2026

Clawdify provides day-zero support for the latest OpenClaw release — now with advanced memory/wiki integration, intelligent context plugins, and enterprise-grade provider support.

If you're running your agents on Clawdify, you can upgrade to v2026.4.11 instantly — no installs, no configs, no downtime.

OpenClaw v2026.4.9 / 4.10 / 4.11 Highlights:

This trio of releases represents a major evolution in agent autonomy, memory, and integration breadth:

🧠 Memory/Wiki Integration with ChatGPT Import — Agents can now import chat histories directly from ChatGPT and compile them into persistent, queryable knowledge bases. The new "Imported Insights" and "Memory Palace" diary features let you inspect and manage imported context directly in the Control UI.
🧩 Active Memory Plugin — A new sub-agent that automatically pulls relevant context before each main reply. Configurable modes (message/recent/full) with live inspection and optional transcript persistence — your agents now have true context awareness.
🤖 Codex Provider Bundle — Native Codex provider with plugin-owned app-server harness, enabling direct thread discovery and model references without manual configuration.
🎬 Video Generation & Asset Management — Enhanced video tools with URL-only asset delivery, typed provider options, audio input references, adaptive aspect-ratio support, and higher image input capacity.
💬 Microsoft Teams & Feishu Expansions — Teams now supports message reactions, listing capabilities, Graph pagination, and delegated OAuth. Feishu gains richer document comment sessions with reactions and typing feedback.
🔌 Plugin Manifests & Exec Policy CLI — Plugins can now declare required auth and configuration steps via manifests. New exec policy CLI for managing tool execution permissions locally.
⚡ Ollama Optimization — Context-window metadata caching during model discovery reduces repeated refetches, speeding up local inference workflows.
🛡️ Security Hardening — Over 30 targeted fixes including browser SSRF defenses, iMessage self-chat detection, Matrix account scoping consistency, WhatsApp account registration, and TTS audio persistence.

Get Started:

Deploy a new agent with OpenClaw 4.11 pre-installed
Upgrade your existing instance in one click from the dashboard — seamless, zero downtime
Import a ChatGPT conversation to build your agent's knowledge base instantly

Your agents stay current — automatically. 🚀

🚀 OpenClaw v2026.4.8 — Already Live on Clawdify

April 7, 2026

Clawdify provides day-zero support for the latest OpenClaw release.

If you're running your agents on Clawdify, you can upgrade to v2026.4.8 instantly — no installs, no configs, no downtime.

OpenClaw v2026.4.7 / 4.8 Highlights:

This is a significant release. OpenClaw 4.7 introduced a wave of new capabilities, and 4.8 followed shortly after with stability fixes for extension loading and an updated release verification process.

What's new:

🔧 Unified inference CLI — openclaw infer provides a single command for model completions, image generation, audio, video, web access, and embeddings. Agents can now handle lightweight and heavy workloads through one interface.
🎵 Music & video editing — agents can now manipulate audio and video files natively.
🔀 Session branching & restore — fork a session, experiment, and roll back if needed.
🔗 Webhook-driven TaskFlows — external systems can now trigger agent workflows via webhooks, opening the door to event-driven automation.
🧠 memory-wiki — a new structured knowledge system with evidence tracking and freshness scoring, replacing flat memory files with persistent, queryable knowledge.
💤 Smarter Dreaming — the dream cycle now learns from session logs in addition to memory files, enabling deeper self-improvement between interactions.
🤖 New model support — Arcee, Gemma 4, and Ollama vision models added to the catalog.

Model Advisory — Claude Opus 4.6:

Community reports suggest Opus 4.6 reasoning depth has degraded in recent weeks, with shorter thinking chains and more frequent execution shortcuts. If you're experiencing issues with agent reliability on Opus, consider switching to GPT-5.4 or Claude Sonnet 4.6 (which remains stable) via your Clawdify model settings. Anthropic's next-generation Claude Mythos is expected soon — early benchmarks show substantial improvements in multi-step agent execution over Opus 4.6.

Clawdify's credit wallet gives you access to 300+ models including GPT-5.4, Sonnet, and Gemini — switch anytime from Settings with no API key management.

Get Started:

Deploy a new agent with OpenClaw 4.8 pre-installed
Upgrade your existing instance in one click from the dashboard — seamless, zero downtime
Switch models instantly via Settings → Paid Models if you're seeing Opus degradation

Your agents stay current — automatically. 🚀

🚀 OpenClaw v2026.4.2 — Already Live on Clawdify

Clawdify provides day-zero support for the latest OpenClaw release.

If you're running your agents on Clawdify, you can upgrade to v2026.4.2 instantly — no installs, no configs, no downtime.

This release represents a massive shift for OpenClaw: moving beyond simple chat and transforming into a true "agent OS" designed for persistent, asynchronous workflows.

Highlights from this release:

📋 Durable Task Orchestration (/tasks): Task Flow is now fully restored as the system backbone. Agents can run long-lived, multi-step workflows with state persistence, revision history, and recovery tooling. Tasks survive restarts and can be resumed or inspected at any time, with support for managed child tasks and async execution.
🔀 Advanced Failover Routing: A new routing layer introduces retry caps, provider-aware failover, and cross-model switching. When APIs fail or rate-limit, agents automatically reroute instead of stalling, enabling graceful degradation under real conditions.
🔒 Strict Secrets Isolation: Credentials are now scoped per agent via auth-profiles.json, eliminating shared global state. This enables safer multi-agent setups and reduces the risk of credential leakage across workflows.
⚙️ Execution & Plugin Hardening: Improved execution defaults reduce friction while tightening plugin activation boundaries and isolating tool usage. Combined with hardened provider transport, this makes execution more predictable under load.
📱 Android Assistant Integration: Launch OpenClaw directly from Android via Google Assistant and pass prompts into running agents, enabling quick, voice-driven task initiation.
🔌 Ecosystem Expansion: Native integrations now include SearXNG search, AWS Bedrock Guardrails, Feishu workflows, and support for newer models like GLM-5.1, Kimi, and Copilot.
🛡️ Bulletproof Background Upgrades (Clawdify Exclusive): We've upgraded our daemon with advanced systemd armor and cache-busting tech. Upgrades now happen seamlessly in the background without aggressive OS interruptions or stuck black screens.

How this hits reality:

This turns OpenClaw from "run one task at a time" into "manage ongoing work." You can now kick off background jobs, switch models mid-session without breaking flow, and rely on automatic failovers when provider APIs go down. For anyone running multi-step agents or production workflows, this removes a huge amount of manual babysitting. Your agents are now persistent, autonomous systems.

Get Started:

Deploy a new agent with OpenClaw 4.2 pre-installed
Upgrade your existing instance in one click from the dashboard — our daemon will automatically migrate your existing API keys to the new secure vaults seamlessly.

Your agents stay current — automatically. 🚀

🚀 OpenClaw v2026.3.31 — Already Live on Clawdify

Clawdify provides day-zero support for the latest OpenClaw release.

If you're running your agents on Clawdify, you can upgrade to v2026.3.31 instantly — no installs, no configs, no downtime.

Highlights from this release:

🔄 Unified Background Tasks: A powerful new SQLite-backed engine to track, manage, and recover cron jobs, subagents, and complex ACP flows.
🛡️ Hardened Security & Approvals: Stricter defaults for dangerous plugin installs, enforced node-pairing checks, and native execution approvals directly inside Slack.
💬 Expanded Channel Integrations: Brand new native QQ Bot support, rich media (video & audio) sending for LINE, and major upgrades to Matrix (live streaming & HTTP proxy routing).
🔌 Advanced MCP Capabilities: Full support for connecting to remote HTTP/SSE Model Context Protocol servers with secure credential handling.
🩹 Core Stability: Dozens of critical fixes, including OpenAI Responses API tool schema normalization, better Anthropic failovers, and Android 15+ pairing improvements.

Both background orchestration and enterprise-grade security see massive upgrades in this version.

Get Started:

Deploy a new agent with OpenClaw 3.31 pre-installed
Upgrade your existing instance in one click from the dashboard

Your agents stay current — automatically. 🚀

🚀 OpenClaw v2026.3.28 — Already Live on Clawdify

Clawdify provides day-zero support for the latest OpenClaw release.

If you're running your agents on Clawdify, you can upgrade to v2026.3.28 instantly — no installs, no configs, no downtime.

Highlights from this release:

🛡️ Plugin Approval Hooks: Any tool execution can pause for explicit user approval
⚡ xAI Responses API + x_search: Expanded model and retrieval capabilities
💬 ACP Bindings: Native integrations for Discord and iMessage
🩹 Stability Fixes: WhatsApp echo loop, Telegram message splitting, Discord reconnect improvements

Both control and system stability see meaningful upgrades in this version.

Get Started:

Deploy a new agent with OpenClaw 3.28 pre-installed
Upgrade your existing instance in one click from the dashboard

Your agents stay current — automatically. 🚀

💰 Credit Wallet & OpenClaw v2026.3.24

March 27, 2026

Clawdify introduces the Credit Wallet — a rechargeable token wallet that gives you instant access to 300+ premium AI models through a single unified provider. We also ship same-day support for OpenClaw v2026.3.24.

Credit Wallet — How It Works:

💳 One-Click Top-Up: Add credits with a single click using your card already on file — no re-entering payment details. Choose from $5, $10 (5% bonus), or $25 (10% bonus) packs.
🤖 openrouter/auto: Our recommended default model. Behind it, a classification layer scores each prompt's complexity against 26 models and routes to the best one at the lowest cost. Simple prompts hit budget models; complex ones route to frontier models automatically.
📊 Real-Time Metering: Every request shows exact token usage and cost in your transaction history. Your balance is always visible in the dashboard header.
🎯 Curated Model Selection: Out of the box you get access to GPT-4o, Claude Sonnet 4, Gemini 2.5 Flash, DeepSeek R1, and more — across budget, mid-tier, premium, and reasoning categories.
🛡️ Spend Protection: Pre-flight cost estimation blocks requests that would exceed your balance. Low-balance warnings notify you before you run out.
🔌 Zero Config: Once you top up, the wallet provider appears automatically in your OpenClaw instance. Pick any wallet model from the model selector and start chatting — no API keys needed.

Getting Started:

Go to Settings in your dashboard
Find the Credit Wallet card
Click any top-up amount — your saved card is charged instantly
Open Chat and select a wallet model like openrouter/auto or openai/gpt-4o
Your balance updates in real-time as you use tokens

OpenClaw v2026.3.24 Highlights:

🔗 OpenAI API Sub-Agents: Talk to sub-agents with @mentions via the improved OpenAI API integration.
🛠️ Skill & Tool Management: New Control UI panels for managing skills and tools directly from the dashboard.
💬 Slack Interactive Replies: Slack channel integration now supports interactive reply buttons.
🏢 Native Microsoft Teams: Full native Microsoft Teams channel support — no third-party bridges required.
🧵 Smart Discord Threads: Discord auto-thread naming brings context-aware conversation organization.

All new Clawdify servers will receive OpenClaw v2026.3.24 automatically (existing servers can be updated with one click).

⬆️ Plan Upgrades, Auto-Updates & OpenClaw v2026.3.23

March 24, 2026

Clawdify now supports seamless plan upgrades, automatic daemon updates, and ships with the latest OpenClaw engine.

What's New in Clawdify:

📐 In-Place Plan Upgrades: Upgrade from Basic → Pro → Max directly from the pricing page. Your server resizes in ~60 seconds with zero data loss — same IP, same subdomain, same config. Stripe handles prorated billing automatically.
🎁 7-Day Free Trial: New users can try the Basic plan free for 7 days. Your card is collected but not charged until the trial ends. Upgrade to Pro or Max during the trial at no cost — the trial carries over to the new plan.
🔄 Automatic Daemon Updates: Your Clawdify daemon now self-updates with cryptographically signed code pushed from the sync API. No more manual OTA pushes — version mismatches are resolved automatically within seconds. Disable auto-updates from Settings if you prefer manual control.
🧾 Trial-Aware Billing: The account page shows your trial status with a countdown, and the billing panel hides $0 invoices during trial periods for a cleaner experience.

OpenClaw v2026.3.23 Highlights:

🔒 Gateway Supervision Fix: Eliminates crash-loop lock conflicts under systemd — fewer random gateway outages on your Clawdify server.
🌐 Gateway Probe Fix: Successful WebSocket handshakes no longer falsely time out as unreachable, improving dashboard connectivity reliability.
💰 OpenRouter Pricing Fix: Stops recursive pricing refresh during bootstrap, so OpenRouter auto routes populate cached pricing correctly again.
🤖 Mistral Safe Defaults: Lowered Mistral max-token defaults to safe output budgets, preventing deterministic 422 rejections on fresh setups.
🔐 Canvas Auth: Canvas routes now require authentication, tightening security on agent-generated web content.

🆓 OpenRouter Integration: Unlimited Free Models Are Here!

March 16, 2026

Clawdify now officially supports OpenRouter as a first-class provider — unlocking access to dozens of high-quality free models including Meta's Llama 3.3 70B, Google's Gemini 2.5 Flash, and Mistral Small 3.1, all with zero API costs.

What's New in Clawdify:

🧙 One-Click Onboarding Wizard: New users are guided through connecting their OpenRouter account in seconds. Pick a free model, set it as your global default, and start chatting — no credit card, no API key, no friction.
🌐 15+ Provider Ecosystem: OpenRouter joins our growing roster of supported providers including OpenAI, Anthropic, Google, Groq, xAI, Mistral, Cerebras, Venice, and more. Switch between them freely from the Terminal's session override dropdown.
🎰 Auto-Route Mode: Select "Auto-Route" and let OpenRouter dynamically pick the best available free model for each request — your agent always gets the fastest, highest-quality response at zero cost.
🔑 Secure OAuth Integration: Connect your OpenRouter account via industry-standard PKCE OAuth flow. No API keys to copy-paste — just click, authorize, and go.
🧪 LLM Terminal: Instantly test any connected model directly from the dashboard. Switch providers and models mid-session with the new interactive override capsule in the terminal header.

🚀 Clawdify Now Supports OpenClaw v2026.3.13

March 14, 2026

We are thrilled to announce day-one support for OpenClaw v2026.3.13! This massive update brings a completely overhauled dashboard experience and blazing-fast model interactions to your Clawdify agents.

What's New in Clawdify:

🎛️ Dashboard v2 Integration: Experience the brand new modular OpenClaw Control UI directly from your Clawdify mission control. Includes a new command palette, mobile bottom tabs, and richer chat tools like slash commands.
⚡ GPT-5.4 Fast Mode: We've enabled configurable session-level fast toggles for OpenAI models. Your agents will now respond faster than ever with optimized request shaping.
🛡️ Hardened Security: We've implemented the latest Gateway/WebSocket browser origin validation, ensuring your agent's cross-site connection remains absolutely impenetrable.
📱 Mobile Refinements: Enjoy the redesigned mobile chat settings UI, making it easier to manage your agents on the go.

🧠 Memory Unleashed: OpenClaw v2026.3.7 & v2026.3.8 Support

March 8, 2026

Clawdify has officially integrated the most intensive OpenClaw update in history! Your agents just got a massive brain upgrade with the new ContextEngine.

What's New in Clawdify:

🔌 Pluggable Context Engine: Agent memory is now completely modular! The new ContextEngine plugin interface means your agent's memory strategies can be customized and swapped seamlessly.
🤖 Dual-Engine Launch: Full native compatibility with OpenAI's GPT-5.4 and Google's Gemini 3.1 Flash.
🔀 Smart Model Routing: When your primary model is overloaded or rate-limited, your Clawdify agent will automatically and seamlessly route to a fallback model instead of failing. Uninterrupted agent uptime!
🛠️ Stability Sweep: Over 200 under-the-hood bug fixes making your autonomous deployments more reliable than ever.

📄 PDF Analysis & Voice Updates: OpenClaw v2026.3.2

March 3, 2026

Your Clawdify agents can now read, listen, and manage secrets better than ever with our rollout of OpenClaw v2026.3.2 support.

What's New in Clawdify:

📚 Native PDF Analysis: Just drop a PDF into the chat. Agents can now natively parse and analyze PDFs using Anthropic and Google as processing backends, complete with extraction fallback strategies.
🎙️ Speech-to-Text (STT) API: Agents can now process audio files directly via the new STT integrations.
🔐 SecretRef Upgrades: The credential referencing mechanism has been expanded to 64 targets, securing the entire planning and execution process.
💬 Telegram Enhancements: Telegram message streaming now defaults to "partial" mode, giving you lightning-fast real-time typing previews when interacting with your agent via Telegram.