mkdocs.yml
Rama main · réplica de lectura del repositorio original.
site_name: DeltaLLM Documentation
site_description: Open-source LLM gateway with unified API, RBAC, guardrails, and observability
# Preserve the language/version path used by Read the Docs in canonical URLs and the sitemap.
site_url: !ENV [READTHEDOCS_CANONICAL_URL, 'https://docs.deltallm.io/en/latest/']
repo_url: https://github.com/deltawi/deltallm
repo_name: deltallm
# Internal plans live beside the public sources for repository-local review, but
# they must never be rendered, indexed, or included in the public sitemap.
exclude_docs: |
internal/
# Shared snippets are pulled into pages by pymdownx.snippets, not published on their own.
includes/
theme:
name: material
# Reuse the UI's built-in assets so the product and docs have one brand source.
custom_dir: ui/public
logo: brand/deltallm-delta.svg
favicon: favicon.svg
palette:
- scheme: default
primary: custom
accent: custom
toggle:
icon: material/brightness-7
name: Switch to dark mode
- scheme: slate
primary: custom
accent: custom
toggle:
icon: material/brightness-4
name: Switch to light mode
features:
- navigation.sections
- navigation.indexes
- navigation.path
- navigation.instant
- navigation.instant.progress
- navigation.tracking
- navigation.top
- navigation.footer
- toc.follow
- search.suggest
- search.highlight
- search.share
- content.code.copy
- content.tabs.link
markdown_extensions:
- admonition
- pymdownx.details
- pymdownx.superfences
- pymdownx.tabbed:
alternate_style: true
- pymdownx.highlight:
anchor_linenums: true
- pymdownx.inlinehilite
- pymdownx.snippets:
base_path:
- docs/includes
check_paths: true
- tables
- toc:
permalink: true
- attr_list
- def_list
extra_css:
- stylesheets/brand.css
nav:
- Home: index.md
- Get started:
- getting-started/index.md
- Start with Docker: getting-started/docker.md
- Check or add your first model: getting-started/first-model.md
- Send your first request: getting-started/quickstart.md
- Create an application key: getting-started/first-api-key.md
- Set up a development environment: getting-started/installation.md
- Video tutorials: tutorials/index.md
- Build with DeltaLLM:
- features/index.md
- Models and providers:
- Choose and connect a model: guides/models-and-providers.md
- Routing and reliability:
- Route traffic and handle failures: guides/routing-and-failover.md
- Cache repeated requests: features/caching.md
- Process work in batches: guides/batching.md
- Realtime audio over WebSockets: guides/realtime.md
- Tools and safety:
- Connect your first MCP server: getting-started/mcp-quickstart.md
- Add your first guardrail: guides/guardrails.md
- Build complete workflows: guides/build-workflows.md
- Use coding agents: guides/coding-agents.md
- Administer DeltaLLM:
- admin-ui/index.md
- Start administering:
- Sign in and get access: admin-ui/access-requirements.md
- Configure the gateway:
- Add and manage models: guides/admin-models.md
- Reuse provider credentials: guides/admin-provider-credentials.md
- Manage route groups: guides/admin-route-groups.md
- Prompts: admin-ui/prompt-registry.md
- MCP servers, bundles and approvals: admin-ui/mcp.md
- Connected accounts: admin-ui/connected-accounts.md
- Test a model: admin-ui/playground.md
- Manage organizations and access:
- Organizations: admin-ui/organizations.md
- Teams: admin-ui/teams.md
- People and permissions: admin-ui/people-and-access.md
- Application keys: admin-ui/api-keys.md
- Coding agents: admin-ui/coding-agents.md
- Plans and tiers: guides/admin-tiers.md
- Set up sign-in and SSO: guides/admin-authentication.md
- External customer sign-in: configuration/external-customer-auth.md
- Connect a Console: guides/console-gateway-connection.md
- Monitor and control:
- Dashboard: admin-ui/dashboard.md
- Usage and spending: admin-ui/usage.md
- Set budgets: guides/admin-budgets.md
- Set rate limits: guides/admin-rate-limits.md
- Guardrails: admin-ui/guardrails.md
- Audit logs: admin-ui/audit-logs.md
- Run operations:
- Administration workflows: guides/admin-workflows.md
- Batch jobs: admin-ui/batch-jobs.md
- Settings: admin-ui/settings.md
- Remove an organization: features/organization-deletion.md
- Deploy and operate:
- deployment/index.md
- Choose where to run DeltaLLM:
- Railway: deployment/railway.md
- Docker: deployment/docker.md
- Kubernetes: guides/kubernetes-deployment.md
- Prepare for production:
- Choose what to turn on: deployment/feature-switches.md
- Production checklist: deployment/production-checklist.md
- Size production for RPS targets: deployment/production-rps-requirements.md
- Secure the deployment: security/hardening.md
- Tune provider connections: guides/provider-connections.md
- Monitor the system: guides/monitoring.md
- Maintain the system:
- Upgrade or roll back: guides/deployment-workflow.md
- Move from LiteLLM: guides/migrate-from-litellm.md
- Check a deployment: deployment/live-smoke-suite.md
- Update the database: deployment/database-migrations.md
- Recover the v0.1.48 model migration: deployment/model-api-identity-migration-recovery.md
- Back up and restore data: deployment/backup-and-restore.md
- Respond to incidents: deployment/incident-runbooks.md
- Understand the system:
- concepts/index.md
- Architecture: concepts/architecture.md
- What happens to a request: concepts/request-lifecycle.md
- Accounts, teams, and access: concepts/tenancy-and-access.md
- Reference:
- reference/index.md
- API:
- Overview: api/index.md
- Common API behavior: api/conventions.md
- OpenAPI file: api/openapi.md
- Model and file endpoints: api/proxy.md
- OpenAI SDK compatibility: api/openai-sdk-compatibility.md
- OpenAI Responses: api/responses.md
- Anthropic Messages: api/messages.md
- MCP endpoints: api/mcp.md
- Administration endpoints: api/admin.md
- Health and metrics: api/health.md
- Configuration:
- Overview: configuration/index.md
- Model deployments: configuration/models.md
- Request shaping: configuration/request-shaping.md
- Routing settings: configuration/router.md
- Model router administration: configuration/model-router-admin.md
- General settings guide: configuration/general.md
- All settings: configuration/general-settings-reference.md
- Providers:
- Overview: providers/index.md
- Models and features: providers/capabilities.md
- Compatible chat providers: providers/compatible-chat.md
- Mistral AI: providers/mistral.md
- SGLang: providers/sglang.md
- OpenAI-compatible: providers/openai-compatible.md
- Feature details:
- Routing and failover: features/routing.md
- Batch processing: features/batching.md
- Caching: reference/caching.md
- MCP tools: features/mcp.md
- Guardrails: features/guardrails.md
- Administration details:
- Access and permission matrix: reference/admin-access.md
- Model administration: admin-ui/models.md
- Plans and tiers: admin-ui/tiers.md
- Provider credentials: admin-ui/named-credentials.md
- Route-group administration: admin-ui/route-groups.md
- Authentication and SSO: features/authentication.md
- Supplied API key design: design/supplied-api-keys.md
- Budgets and spending: features/budgets.md
- Rate limiting: features/rate-limiting.md
- Audit records: features/audit-log.md
- Operations details:
- Kubernetes and Helm: deployment/kubernetes.md
- Metrics and callbacks: features/observability.md
- Alert destinations: features/alerting.md
- Measure gateway concurrency: deployment/concurrency-measurement.md
- Gateway ingress admission: deployment/ingress-admission.md
- Dependency capacity: deployment/dependency-capacity.md
- Database pooler (PgBouncer): deployment/database-pooler.md
- Concurrency runtime design: design/concurrency-runtime.md
- Process lifecycle: deployment/process-lifecycle.md
- Saturation autoscaling: deployment/saturation-autoscaling.md
- Capacity profiles: deployment/capacity-profiles.md
- Budget and prompt performance: deployment/budget-prompt-performance.md
- Spend recovery: deployment/spend-recovery.md
- Accounting protocol v2: deployment/accounting-v2.md
- Accounting v2 acceptance and soak: deployment/accounting-v2-acceptance.md
- Accounting protocol v2 design: design/accounting-protocol-v2.md
- Bounded in-memory state: deployment/bounded-state.md
- Request deadlines: deployment/request-deadlines.md
- Provider connection tuning: deployment/upstream-http.md
- Provider connection pool design: design/upstream-http-pool.md
- Selector session affinity design: design/selector-session-affinity.md
- Upgrade and rollback behavior: deployment/upgrade-and-rollback.md
- Release-specific rollouts: deployment/rollout-index.md
- Governance migration: deployment/governance.md
- Organization tiers rollout: deployment/organization-tiers-rollout.md
- Batch scheduler rollout: deployment/batch-scheduler-rollout.md
- Batch webhook rollout: deployment/batch-webhook-rollout.md
- Router state cutover: deployment/router-state-schema-cutover.md
- Telemetry ingestion rollout: deployment/telemetry-ingestion-rollout.md
- Usage reporting rollout: deployment/usage-reporting-v2.md
- Performance evidence:
- RPS results: project/issue-320-rps-report.md
- Reproduce RPS tests: project/issue-320-rps-reproduction.md
- Concrete concurrency measures: project/issue-320-50-to-500-rps-measures-and-integration.md
- Policies and support:
- Security model: security/index.md
- Report a security problem: security/vulnerability-reporting.md
- Verify release images: security/release-verification.md
- Versions and compatibility: project/versioning-and-compatibility.md
- Documentation policy: project/documentation-governance.md
extra:
social:
- icon: fontawesome/brands/github
link: https://github.com/deltawi/deltallm
- icon: fontawesome/brands/youtube
link: https://www.youtube.com/@deltallm
name: DeltaLLM on YouTube