Open source alternatives to fal
200 open source apps like fal, closest match first.
Top picks
- F

Falcon LLM
LaunchFalcon LLM is a generative large language model (LLM) that helps advance applications and use cases to future-proof our world.
Open source
- L

Llamafarm
LaunchDeploy any AI model and related databases, RAG, agents, and pipelines to any device in 30 seconds. No cloud required.
FreeOpen source
- A
Airunner
LaunchOffline inference engine for art, real-time voice conversations, LLM powered chatbots and automated workflows
FreeOpen sourceSelf-hosted
- G
GoModel
LaunchAI gateway / AI control plane / AI proxy written in Go. Unified OpenAI-compatible and Anthropic-compatible API for OpenAI, Anthropic, Gemini, Groq, xAI, Ollama, vLLM and more. A LiteLLM alternative with observability, guardrails, streaming,
FreeOpen sourceSelf-hosted
- A

AIClient2API
LaunchSelf-hosted multi-protocol AI API proxy for Antigravity, Codex, Grok, Kiro, OpenAI, Claude, and custom providers. Supports OpenAI-compatible API, Claude API, Gemini protocol conversion, GPT, Grok Build, Claude Opus, Gemini Pro, Kimi, MiniMa
FreeOpen sourceSelf-hosted
- A

AI-Flow
LaunchAI-Flow is an open-source web app for building workflows that chain AI models together. You connect models from OpenAI, Anthropic, Google and Replicate into one flow, so the output of one model can feed the next, for example an image model
Open source
More like fal
OmniKey-AI-Unified-Key-Manager
A self-hosted LLM proxy and failover gateway supporting OpenAI & Gemini…
About fal
fal
Build real-time AI applications with lightning fast inference (under 120ms)
fal is a platform for running generative media models, meaning image, video, audio and 3D models, through one API. Its Model APIs give you access to more than 1,500 models, so you can call them without setting up separate providers, and the site says new models are available as soon as they launch. Its Serverless product lets you deploy your own custom models and run them on autoscaling GPUs, with built-in observability and performance controls for latency and throughput. There is also a Compute offer for renting GPUs directly, including RTX PRO 6000, H100, H200, B200, GB200 and B300 cards, with list and lower starting prices shown on the pricing page. Video models are billed per second or per output video, and the pricing page lists units per dollar for each model. The site also offers enterprise plans with security and compliance, documentation, a blog and a GenMedia report.
- Price
- Paid
- Category
- AI
- Website
- fal.ai


