Claude Code OAuth Login Timeout Bug on Windows
Claude Code, an AI coding assistant from Anthropic, has a confirmed bug preventing Windows users from logging in. The issue affects version 2.1.92 and causes complete login failure through Google OAuth authentication.
Bug Details
The bug manifests as an OAuth timeout error when attempting to log in on Windows systems. Users report the exact error message: OAuth error: timeout of 15000ms exceeded. This occurs consistently regardless of login attempts, effectively blocking all access to Claude Code on affected Windows machines.
Reproduction Steps
- Open Claude Code on Windows
- Start login using Google authentication
- Complete the sign-in flow in the browser
- Return to Claude Code
- Receive the OAuth timeout error
The issue reporter confirmed using Claude Code version 2.1.92 on Windows with WSL (Windows Subsystem for Linux). The bug was reported on April 6, 2026, and has generated significant discussion with 56 points and 35 comments on Hacker News, indicating it affects multiple users.
Impact
This is a blocking bug that prevents Windows users from accessing Claude Code entirely. The 15-second timeout occurs during the OAuth handshake process, suggesting either network configuration issues, firewall interference, or a bug in the authentication flow implementation specific to Windows environments.
For developers relying on Claude Code for AI-assisted programming, this represents a significant workflow disruption until Anthropic releases a fix. The issue is labeled as a bug in the GitHub repository but currently has no assignee or milestone for resolution.
📖 Read the full source: HN AI Agents
👀 See Also
M5 Max vs M3 Max Inference Benchmarks for Qwen Models on oMLX
Benchmarks comparing M5 Max and M3 Max MacBook Pros running Qwen 3.5 models via oMLX v0.2.23 show M5 Max delivering 1.4-1.7x faster token generation and up to 4x faster prefill at long contexts.
Claude Artifacts API Usage Counts Against Chat Quota, Not API Billing
Using Claude artifacts within Claude makes normal API calls that are intercepted by Anthropic and authenticated through the logged-in session, counting against a plan's chat quota rather than API billing. Users can verify this by testing artifacts and checking that API usage remains at zero in the Claude Console.
OpenClaw Client Adds Cost Tracking and Per-Agent Spending Limits
New release adds spending caps per agent, live usage UI with circular progress bar, sub-agent management, skill toggling, and per-agent model selection.
Cannes Film Cost $500k to Make, $400k Was AI Compute Costs
A film screened at Cannes cost $500k to produce; $400k of that was spent on AI compute. A striking data point for AI agent developers building generative video pipelines.