hermes-vault/Brain/preferences/pref-default-model-deepseek.md

kind, id, created_at, _confirmed_at, unconfirmed_until, tags, topic, _status, principle, _evidenced_by, _applied_count, _violated_count, _last_evidence_at, _confidence, _confidence_value, pinned, scope

kind	id	created_at	_confirmed_at	unconfirmed_until	tags	topic	_status	principle	_evidenced_by	_applied_count	_violated_count	_last_evidence_at	_confidence	_confidence_value	pinned	scope
brain-preference	pref-default-model-deepseek	2026-07-12T20:50:36Z	2026-07-12T20:50:36Z	2026-07-12T20:50:36Z	brain brain/preference brain/topic/default-model-deepseek brain/scope/cost-efficiency	default-model-deepseek	confirmed	Nutze als Standard-Modell DeepSeek V4 Flash, nicht Qwen Max. Qwen Max nur auf explizite Anforderung oder für extrem komplexe Architektur-Tasks. DeepSeek kostet ~$4/Monat vs Qwen ~$21/Monat bei gleicher Nutzung.	sig-2026-07-12-default-model-deepseek	0	0	null	low	null	false	cost-efficiency

Origin

• sig-2026-07-12-default-model-deepseek