[BUG] windows 11 home issue

Preflight Checklist

• [x] I have searched existing issues and this hasn't been reported yet
• [x] This is a single bug report (please file separate reports for different bugs)
• [x] I am using the latest version of Claude Code

What's Wrong?

VM starts, network connects, but API reachability shows UNREACHABLE. Error: "Failed to start Claude's workspace - Can't reach the Claude API from Claude's workspace." Previously also got "VM service not running" before enabling Hyper-V on Windows Home via DISM script.

What Should Happen?

Cowork workspace should start successfully and connect to the Claude API, allowing task execution. The VM boots and gets network connectivity (Network status: CONNECTED), but the API reachability check fails (API reachability: UNREACHABLE). Regular Chat mode works perfectly — only Cowork is affected.

If there's a "What Actually Happens?" field:

Error Messages/Logs

Failed to start Claude's workspace
Can't reach the Claude API from Claude's workspace.
Restarting Claude or reconnecting to your network sometimes resolves this

Steps to Reproduce

Install Claude Desktop v1.1.2685.0 on Windows 11 Home 25H2 (Build 26200.7840)
Sign in with a paid plan (Max plan)
Enable Windows features: Virtual Machine Platform, Windows Hypervisor Platform, Windows Subsystem for Linux
Install WSL2 via wsl --install (Ubuntu [distribution)](url)
Enable Hyper-V on Windows Home using DISM package installation script (required since Windows Home lacks native Hyper-V)
Restart computer
Open Claude Desktop → Click "Cowork" tab
Wait for VM to download and initialize (~3-5 minutes first time)
VM boots successfully, network shows CONNECTED
API reachability check fails: PROBABLY_UNREACHABLE → UNREACHABLE
Error displayed: "Failed to start Claude's workspace - Can't reach the Claude API from Claude's workspace"
Attempted fixes that did NOT resolve: disabling all firewalls, adding outbound firewall rule for Claude, flushing DNS, resetting winsock, setting Google DNS (8.8.8.8) in WSL, deleting and re-downloading VM bundle, running Claude as administrator
Note: At one point a task DID spawn and run for ~6 minutes (process exited code=0) despite the API unreachable warning, but this was not consistent
Chat mode works perfectly — only Cowork is [affected[](url)](url)

Claude Model

Opus

Is this a regression?

No, this never worked

Last Working Version

_No response_

Claude Code Version

claude for windows version 1.12685 (f39a62)

Platform

Anthropic API

Operating System

Windows

Terminal/Shell

Windows Terminal

Additional Information

If it still fails — Cem, I'll be direct. We've done everything possible on your end: enabled all VM features, installed Hyper-V on Home, configured WSL2, set DNS, created firewall rules, reset the VM bundle. Your system is properly set up.
This is Anthropic's bug to fix. The VM boots, gets network, but can't route to api.anthropic.com. It's been reported by multiple Windows users since the Feb 10 launch.
I'd recommend doing two things now:

Click "share your debug logs" in the Cowork error banner — this sends your logs directly to Anthropic
Go to github.com/anthropics/claude-code/issues/24918 and add a comment with your specs (Windows 11 Home, i5-10300H, 8GB RAM) — the more reports, the faster they'll patch it.Windows 11 Home 25H2 (Build 26200.7840)
Intel Core i5-10300H, 8GB RAM, x64
Claude Desktop v1.1.2685.0
WSL2 Ubuntu installed, Virtual Machine Platform ✅, Windows Hypervisor Platform ✅, Hyper-V enabled via DISM workaround
No VPN, no third-party antivirus
Firewall tested both on and off — same result

cowork_vm_node.log