Allow a global default-model override for the openrouter backend #76

New issue

Open

opened 2026-10-15 19:15:59 +00:00 by agent-pi · 0 comments

agent-pi commented

2026-10-15 19:15:59 +00:00

Collaborator

The openrouter backend has a fixed default model (deepseek/deepseek-v4.1-flash) baked into the preset. Per-agent model selection covers per-run overrides (agents: code:vendor/model), but an operator who wants a different default for all agents without listing every agent explicitly has no knob today — they must either spell out every agent with name:model or accept the preset default.

Found during the review of #75 (the openrouter_key input description documents the default inline because there is no separate setting).

A parallel observation: the generic openai backend has WUMING_OPENAI_MODEL for exactly this, while the hosted presets do not have an equivalent.

The `openrouter` backend has a fixed default model (`deepseek/deepseek-v4.1-flash`) baked into the preset. Per-agent model selection covers per-run overrides (`agents: code:vendor/model`), but an operator who wants a different default for **all** agents without listing every agent explicitly has no knob today — they must either spell out every agent with `name:model` or accept the preset default. Found during the review of #75 (the `openrouter_key` input description documents the default inline because there is no separate setting). A parallel observation: the generic `openai` backend has `WUMING_OPENAI_MODEL` for exactly this, while the hosted presets do not have an equivalent.

agent-pi referenced this issue

2026-10-15 19:16:26 +00:00

Review this repo's PRs via OpenRouter in a single step #75

agent-pi added the

Kind/Feature

Priority

Low

labels

2026-10-15 19:46:23 +00:00