ProductFeaturesDocs
Built InUI ComponentsAdmin ConsoleDatabaseBackendStorageAIRAGAuthSecurity
CompanyStoryAboutPricingBlogRoadmap
Install FLIN
Built into the language

AI-Native from Day One

Not bolted on. Built in. Semantic search, embeddings, RAG, and multi-provider AI — all native FLIN functions. No libraries, no SDKs, no API wrappers.

AI as a Language Feature

Other frameworks bolt AI on. FLIN bakes it in. Six AI capabilities, zero configuration.

Semantic Text Type

Add semantic before any text field and FLIN auto-generates embeddings on save. One keyword makes your data AI-searchable.

Vector Search

search "query" in Entity — find records by meaning, not keywords. Returns results ranked by semantic similarity, not string matching.

RAG

Retrieval-Augmented Generation. Semantic search finds relevant documents, injects them as context, and streams an AI answer — all in native FLIN code.

ask_ai()

Call any LLM provider with one function. Pass a prompt, get an answer. Provider, model, and parameters — all configurable, all optional.

AI Gateway

Built into the admin console at /_flin. Monitor provider stats, check API key configuration, and test prompts — all in one dashboard.

Keyword + Semantic

Side-by-side comparison of keyword and semantic search. Best of both worlds — exact matches when you need them, meaning-based results when you don't.

See It in Action

From data model to AI-powered search in 15 lines of FLIN.

app/knowledge.flin
// 1. Mark a field as semantic — embeddings auto-generated on save
entity Document {
    title: text @required
    content: semantic text         // ← This is the magic
    category: text
}

// 2. Search by meaning — not keywords
results = search "how to configure SSL certificates" in Document

// 3. Ask any LLM — one function, 8 providers
answer = ask_ai("Summarize these search results", [
    "provider": "claude",
    "model": "claude-sonnet-4-20250514"
])

// 4. Or use keyword search for exact matches
exact = keyword_search("SSL", "Document", "content", 10)

Auto-Embeddings

Every time you save an entity with a semantic text field, FLIN generates and indexes embeddings automatically.

8 Providers

OpenAI, Claude, Gemini, Mistral, DeepSeek, Groq, Ollama, OpenRouter. Switch providers with one config change.

Local or Cloud

Use Ollama for fully local embeddings and inference, or connect to any cloud provider. Your data, your choice.

8 AI Providers. One Function.

ask_ai() supports every major LLM provider. Switch between them without changing your application code.

OpenAI

GPT-4o, GPT-4o Mini, o1, o3

Embeddings + Chat

Claude

Claude Opus, Sonnet, Haiku

Chat

Gemini

Gemini 2.0 Flash, Pro

Chat

Mistral

Mistral Large, Medium, Small

Chat

DeepSeek

DeepSeek V3, R1

Chat

Groq

LLaMA, Mixtral — ultra-fast

Chat

Ollama

Any local model — fully offline

Embeddings + Chat

OpenRouter

200+ models via one API key

Chat

RAG Without the Plumbing

In other frameworks, RAG means stitching together a vector database, an embedding API, a chunking strategy, and an LLM client. In FLIN, it's a page.

app/chat.flin
// RAG in FLIN: search → context → stream

entity KnowledgeBase {
    title: text @required
    content: semantic text
}

query = ""

fn askWithContext() {
    // Find the 5 most relevant documents
    docs = search query in KnowledgeBase

    // Ask the AI with those documents as context
    answer = ask_ai(query, [
        "provider": "claude",
        "context": docs
    ])
}

<form submit={askWithContext()}>
    <input bind={query} placeholder="Ask anything..." />
    <button>Ask</button>
</form>

Build AI-Powered Apps Today

Add semantic text to any field. Call ask_ai() from any page. Ship AI features in minutes, not weeks.

curl -fsSL https://flin.sh | bash
Install FLIN AI Documentation