IT Performance

A local model taking two to three minutes to answer "introduce yourself in one sentence" is not normal, even on a mid-range GPU. On my AMD Radeon RX 6800 XT the cause turned out to be two stacked problems: the agent app was sending a ~20,000-token prompt with every message, and Ollama had silently fallen back from ROCm to Vulkan, where long prompts are processed very slowly. Fixing the second one took a 25,000-token prompt from about 9 minutes to about 30 seconds. Everything below was measured on one Windows 11 machine on 2026-09-28 with Ollama 0.34.4 and qwen3:8b / qwen3:14b (Q4_K_M). Your numbers will differ; the diagnostic steps will not. The symptom A desktop AI agent pointed at local Ollama showed "Worked for 2m 53s" for a one-line greeting. Cloud models in the same app answered in seconds, so the app itself looked fine. Short prompts typed directly into ollama run felt fast, which made the slowness look random. Before: a one-line greeting took 2 ...

Your Claude Code session eats through the 5-hour limit before lunch. Or the opposite: answers come back thin on a problem you know is hard. Both usually trace to one setting most people never open — effort . The symptom The usage meter (5h / 7d) climbs much faster than it used to, with no change in how much you type. Or: responses arrive fast but shallow on genuinely hard debugging, and you find yourself re-prompting three times. Switching models didn't fix it, and you can't find a "verbosity" or "quality" slider in the docs you read. The cause: effort is a per-turn reasoning budget Claude Code exposes a setting called effort that controls how much reasoning the model spends on each turn. Run /effort and you get a slider labelled Faster on the left and Smarter on the right, with five stops: low  ·  medium  ·  high  ·  xhigh  ·  max These are the descriptions the product itself attaches to each stop: Level What it means low Q...