
Ollama “pull model manifest: 412”: The Simple Fix, and Why Upgrading Is Not Always Enough
Ollama shows "Error: pull model manifest: 412" when the model you are pulling needs a newer Ollama than the one you run. Upgrade the official way,…
AI, hardware, and the tools I actually use. The notes are all here.

Ollama shows "Error: pull model manifest: 412" when the model you are pulling needs a newer Ollama than the one you run. Upgrade the official way,…

Ollama shows "error loading model: unknown model architecture" when the model file uses an architecture its bundled llama.cpp does not know yet. Upgrade Ollama, use the…

Claude Code API Error 401 OAuth token has expired means the login token expired and the automatic refresh failed. We reproduced it on Claude Code 2.1.312…

"Codex could not start its user interface" appeared for many VS Code users after the Codex extension updated to 26.1021.22027. The GitHub issue is still open…

Claude Code Prompt is too long means the request, including system prompt, tools and CLAUDE.md, is bigger than the model's context window. We reproduced the variants…

The OpenAI API returns 403 "Country, region, or territory not supported" when a request comes from a country or territory that is not on OpenAI's supported…

OpenAI returns 401 "Incorrect API key provided" when the key your request sends is wrong: a typo, a deleted key, or a key from another organization…

Claude Code Unable to connect to API means the connection to the API never completed. Usually the network is fine and a leftover ANTHROPIC_BASE_URL, a dead…

llama.cpp shows "error: --model or -hf is required" when llama-cli starts without a model. Give it a local GGUF with -m or a Hugging Face repo…

When you hit the Cursor Pro usage limit, nothing is charged until you act. Pro is now a legacy plan with two monthly pools, and on-demand…

A leaked API key costs money for as long as it stays active. Delete or disable it first, make a new one, then check what it…

The Artificial Analysis Intelligence Index is one score from nine independent tests. Version 5 (v5.0.1) splits them into four categories at 25% each. Claude Fable 5.2…