Ollama Diagnostics
USB AI DRIVE  /  E:\  _
[!] WRONG CONTEXT — FETCH IS BLOCKED
This page is running inside an Electron / sandboxed app which blocks all fetch() requests to localhost. Tests cannot run here.

Open this file in a real browser instead:
  1. Press Win + R → paste E:\Utilities\OLLAMA_DIAGNOSTICS.html → Enter
  2. Or drag the file into a Chrome or Edge window
  3. Or run E:\Utilities\start_webserver.py then open http://localhost:8080
IDLE
TEST 01 Server Running
WAITING
Tests if Ollama is reachable at http://localhost:11434
TEST 02 List Models
WAITING
Lists all models loaded from E:\Models\
TEST 03 Generate — Simple Message
WAITING
Sends a short test prompt to the selected model. First load may take 30–90 seconds.
[!] CORS / FETCH BLOCK DETECTED — FIXES
OPTION A — ENV VAR

Stop Ollama and restart it with CORS allowed. Open a Command Prompt and run:

set OLLAMA_ORIGINS=*
set OLLAMA_MODELS=E:\Models
E:\ollama\ollama.exe serve
OPTION B — WEBSERVER

Use the Python webserver already on the USB. This avoids all file:// CORS issues.

Run: E:\Utilities\start_webserver.py

Then open:
http://localhost:8080
OPTION C — BAT FILE

The .bat launcher already sets all required environment variables automatically.

Run:
E:\Utilities\
  Utilities_CHECK_MODELS.bat

Then open your chat page.
System Info
Result Guide
[FAIL] Test 1 Ollama not running. Run Utilities_CHECK_MODELS.bat
[OK] 1, [FAIL] 2 Running but no models. Check OLLAMA_MODELS=E:\Models
[OK] 1&2, [FAIL] 3 Model error. Try phi3:mini or free memory.
[OK] All Ollama is working. Issue is in chat HTML page.
Failed to fetch CORS block. See CORS Fix section or use webserver.
USB Drive Paths
LAUNCHER
E:\AI_LAUNCHER.html
CHAT PAGES
E:\Portal Pages\
UTILITIES
E:\Utilities\
OLLAMA EXE
E:\ollama\ollama.exe
MODELS
E:\Models\
DOCUMENTS
E:\Documents\
WEBSERVER
E:\Utilities\start_webserver.py