Tale is a sovereign AI platform: chat with models over your own documents, build custom agents, run multi-step automations, and manage customer conversations — with your choice of AI providers and your data pinned to a region you control. The same platform ships in two editions — Cloud, managed by us in Switzerland and the EU, and Self-hosted, which you run on your own infrastructure. Every feature, every API, every role is identical between the two; only the operational surface differs.Documentation Index
Fetch the complete documentation index at: https://docs.tale.dev/llms.txt
Use this file to discover all available pages before exploring further.
Choose your edition
Start here if you are deciding how Tale is hosted:- Cloud — we run the stack, keep it patched, and host your data in Switzerland or the EU. Best when you want Tale’s capabilities without operating the infrastructure.
- Self-hosted — you run the Docker Compose bundle on your own VPC, hardware, or air-gapped environment. Best when data sovereignty, custom networking, or a custom build is a requirement.
Choose your role
Tale has six roles — Owner, Admin, Developer, Editor, Member, Disabled — each with a distinct permission set. Role-indexed documentation lives under Platform and applies to both editions:- Member — read-only end user.
- Editor — content management and approvals.
- Developer — agents, automations, integrations.
- Admin — organisation settings.
What makes Tale different
- Your data, your region. Cloud pins every tenant to Switzerland or the EU, with explicit subprocessor disclosure. Self-hosted leaves the network entirely, including prompts and document content.
- Any model. OpenAI, Anthropic, Google, Mistral, Meta, or self-hosted models via Ollama, vLLM, or LocalAI — switchable per-agent without migration.
- Built for teams. Six roles from read-only Member up to full Owner, organised into teams with scoped knowledge and access.
- Certified. ISO 27001 and SOC 2 Type II, Swiss-based, GDPR-compliant by default.