Route Private AI Workloads to Local Models with AISBF

AISBF tutorial

Route Private AI Workloads to Local Models with AISBF

Keep sensitive prompts on local or self-hosted backends while still using hosted providers for general work.

Providers dashboard with hosted, OAuth, and local/broker-backed providers.

1. Start local backend

ollama pull llama3.1
ollama serve
curl http://localhost:11434/api/tags

2. Add local provider

Add an Ollama/OpenAI-compatible provider named private-local.

3. Call private provider

curl -X POST -H "Authorization: Bearer $AISBF_API_TOKEN" -H "Content-Type: application/json" \
  -d '{"model":"user-provider/private-local/llama3.1","messages":[{"role":"user","content":"Summarize this internal incident log."}]}' \
  "https://aisbf.cloud/api/u/$AISBF_USERNAME/chat/completions"

Privacy checklist

Know where AISBF runs.
Know where each backend runs.
Restrict prompt logging.
Self-host when data-path control is mandatory.

Try AISBF

AISBF is open source and also available as a hosted service. During the current testing period, hosted Pro is available as unlimited access for €6/month or €60/year; subscribing helps fund continued AISBF development and infrastructure while this one-human project is still demo-stage.

Support the €6 / €60 Pro test plan View source code