Verified reference stack
Legal — Private RAG — RTX 4070 Ti SUPER
Qwen3-14B package size is about 9.3 GB in Ollama; RAG context must be bounded to preserve VRAM margin.
Open WebUI
Open tool profile →Ollama
Open runtime profile →Qwen3-14B
Open model profile →Fully local RAG with local embeddings/model providers
Apache-2.0
Evidence
Continue the decision
Put this reference stack in context
A reference configuration is most useful when you compare it with alternatives and then apply your own hard constraints.
Private AI for Lawyers & Legal
See the professional workflows and verified baselines connected to this stack.
Self-hosted RAG
Understand the architecture and privacy assumptions behind this use case.
Compare related stacks
Review this configuration beside related verified references.
Run the Stack Finder
Apply your exact profession, task, privacy requirement and hardware.
Freshness rule: Material changes set this stack to REVALIDATION_REQUIRED.