500 model requires more system memory (14.4 GiB) than is available (12.3 GiB)

I was using llama codestral model in the background.

OS VERSION:
Distributor ID: Ubuntu
Description: Ubuntu 24.04 LTS
Release: 24.04
Codename: noble

AI MODEL
NAME ID SIZE MODIFIED
codestral:latest XXXX 12 GB

version: micro-agent v0.1.4

Exception:
at APIError.generate (file:///usr/lib/node_modules/@builder.io/micro-agent/node_modules/openai/error.mjs:62:20)
at OpenAI.makeStatusError (file:///usr/lib/node_modules/@builder.io/micro-agent/node_modules/openai/core.mjs:268:25)
at OpenAI.makeRequest (file:///usr/lib/node_modules/@builder.io/micro-agent/node_modules/openai/core.mjs:311:30)
at process.processTicksAndRejections (node:internal/process/task_queues:95:5)
at async getSimpleCompletion (file:///usr/lib/node_modules/@builder.io/micro-agent/dist/cli.mjs:1055:22)
at async getFileSuggestion (file:///usr/lib/node_modules/@builder.io/micro-agent/dist/cli.mjs:954:9)
at async interactiveMode (file:///usr/lib/node_modules/@builder.io/micro-agent/dist/cli.mjs:735:33)
at async file:///usr/lib/node_modules/@builder.io/micro-agent/dist/cli.mjs:1776:11

Reproduce the failure with the codestral model on Ubuntu 24.04, then inspect the dist/cli.mjs entry points getSimpleCompletion, getFileSuggestion, and interactiveMode shown in the stack trace. Done means the insufficient-memory case is handled or documented with a verified outcome for the reported environment.

Written by the indexing model from the issue text.