Will Aider warn about unknown model settings?

It may, because it has no built-in metadata for these ids. The warning is harmless; Aider falls back to generic settings.

Which model works best for code?

gpt-oss 20B is the best fit for agentic coding in the catalog; GLM-4 32B is strong at structured edits.

Integration · Coding assistant

Pair-program in your terminal with coding models running on your GPUs.

Aider talks to OpenAI-compatible servers when the model name is prefixed with openai/.

Shell

export OPENAI_API_BASE=https://gpuos.si/v1
export OPENAI_API_KEY=$GPUOS_API_KEY
aider --model openai/gpt-oss-20b

For larger refactors, deploy Qwen3 32B or GLM-4 32B on a 24 GB GPU and switch with --model openai/qwen3-32b.

Official documentation: aider.chat

Will Aider warn about unknown model settings?: It may, because it has no built-in metadata for these ids. The warning is harmless; Aider falls back to generic settings.
Which model works best for code?: gpt-oss 20B is the best fit for agentic coding in the catalog; GLM-4 32B is strong at structured edits.

Free for one GPU. Connect a machine, deploy a model, then paste the base URL and your key.