Stop stitching together 12 libraries. Vorte gives you auth, database, cache, queues, AI agents, payments, search, and 14 more modules — all wired up, production-ready, from day one.
⚡Built on FastAPI🦀Rust native engine🐍Python 3.10+TSTypeScript codegen⚖MIT License
21 Built-in Modules
Everything you need. Nothing you don't.
Every module ships with sensible defaults, full TypeScript types, and zero-config integration. Spend time on your product, not on infrastructure glue.
AI-First Architecture
Native multi-provider AI integration (OpenAI, Anthropic, Gemini, Mistral) with automatic cost tracking, streaming responses, embeddings, and intelligent model routing. Switch providers with one line.
21 Built-in Modules
Auth, Database, Cache, Queue, AI, Agents, Storage, Search, Mailer, Payments, Webhooks, GraphQL, Sockets, Scheduler, Feature Flags, Audit Logs, and more — all included, all integrated.
Rust-Native Engine
Optional zero-copy radix tree router, work-stealing async executor, priority-based scheduler, and multi-format serialization (JSON, MessagePack, CBOR). Up to 5× faster than pure Python.
Admin Dashboard
Real-time Next.js dashboard at /_vorte/dashboard — live metrics, per-module health, route inspector, queue viewer, AI cost breakdown, and WebSocket connections monitor.
Production Ready
Kubernetes liveness/readiness probes, Prometheus metrics, structured JSON logging, API versioning with deprecation headers, circuit breakers, and graceful drain on shutdown.
WASM Sandbox
Execute untrusted user code or plugins in isolated WebAssembly sandboxes with fuel metering, memory limits, and optional WASI filesystem access. Safe by default.
TypeScript Codegen
Automatically generate a fully-typed TypeScript client from your Python routes. Every endpoint, every request/response model — always in sync, no manual schema upkeep.
4-Layer Cache
L1 in-process memory, L2 Redis, L3 CDN edge, L4 database query cache — all configurable per-route with a single decorator. Automatic invalidation and stampede protection.
AI Agents + RAG
Build autonomous AI agents with tools, long-term memory, Retrieval-Augmented Generation, multi-agent pipelines, input/output guardrails, and structured output parsing.
Quick Start
From pip install to production in minutes
Install, scaffold, and run. Vorte's CLI wires everything together — database migrations, env validation, module config, and a running admin dashboard. No boilerplate required.
# Install Vorte via pip or uv
$pip install vorte
✓ Downloading vorte-1.0.0...
✓ Installing 21 built-in modules...
✓ Linking Rust native engine...
✓ Vorte is ready. Happy hacking!
$
vorte v1.0.0
Python 3.12Rust engine21 modules
1
Install
pip install vorte
2
Scaffold
vorte new my-api --modules ai,auth,db
3
Develop
vorte serve --watch
vorte performance
< 10msCold Start
~120k req/sThroughput
~28 MB baseMemory
21 built-inModules
Rust engine activeAsync I/O
Architecture
Performance without compromise
Vorte's layered architecture keeps Python's developer ergonomics while delegating hot paths to a Rust engine. You get both: fast iteration and production throughput.
Async-First Core
Every module is built on asyncio with structured concurrency. No sync/async friction, no blocking calls.
Zero-Config ORM
SQLAlchemy async with automatic N+1 detection, query plan logging, and migration generation via Alembic.
Smart Rate Limiting
Token bucket, sliding window, or fixed window — per user, per IP, or per API key. Redis-backed for distributed deployments.
Background Jobs
Priority queues with backpressure, dead-letter queues, automatic retries with exponential backoff, and cron scheduling.
Developers Love Vorte
Built by engineers, for engineers
Hear from the developers who ship with Vorte every day.
“We replaced Flask + Celery + Redis boilerplate with Vorte and cut our backend setup time from 3 days to 30 minutes. The AI module alone is worth it — streaming, cost tracking, fallback routing all just works.”
Sarah Chen
Staff Engineer, Stealth AI Startup
“The Rust engine is not just marketing. We benchmarked at 118k req/s on a $20/month VPS. FastAPI alone got us 42k. The WASM sandbox let us ship user-defined scripts safely in production without a separate service.”
Marcus Okafor
Backend Lead, DevTooling Co.
“Admin dashboard at /_vorte is insanely useful in staging. Seeing real-time queue depths, AI cost per endpoint, and WebSocket connections without setting up Grafana — that's hours of ops work I didn't have to do.”
Priya Nair
Senior Developer, SaaS Collective
“I moved from Django REST Framework to Vorte for a new project. The 4-layer cache and TypeScript codegen together mean my frontend team can iterate without waiting for me to update API contracts.”
James Oduya
Full-Stack Engineer, Fintech Lab
“The AI agents module is the real gem. Multi-step tool-use, RAG pipeline, guardrails — we built a customer support agent in an afternoon. Switching from OpenAI to Anthropic was literally changing one string.”
Leila Mansouri
ML Engineer, AI Products Studio
“MIT license, no vendor lock-in, self-hostable admin dashboard, and first-class Kubernetes support. Finally a Python framework that takes production seriously from day one instead of leaving it as an exercise for the reader.”
Daniel Park
Platform Architect, Platform Engineering
pip install vorte && vorte serve
Stop configuring. Start shipping.
Vorte is MIT-licensed and free forever. Read the docs, scaffold your first app, or drop into the GitHub discussions.
Vorte is MIT licensed. No usage limits, no paywalled modules, no hidden enterprise tier. Use it in any project — commercial, personal, internal.
MIT LicenseActive Project
Community Edition
100% Free. Forever.
We believe in the power of free, collaborative software. Vorte is built with an uncompromising commitment to open-source excellence. You have absolute freedom to build, scale, and host your platforms on your own terms.
