Use Swama from 4D

Abstract

Swama is a high-performance MLX-based LLM inference engine designed specifically for macOS.

Usage

Instantiate cs.swama.swama in your On Startup database method:

var $swama : cs.swama.swama

If (False)
    $swama:=cs.swama.swama.new()  //default
Else 
    var $homeFolder : 4D.Folder
    $homeFolder:=Folder(fk home folder).folder(".MLX")
    var $file : 4D.File
    var $URL : Text
    var $port : Integer
    
    var $event : cs.event.event
    $event:=cs.event.event.new()
    /*
        Function onError($params : Object; $error : cs.event.error)
        Function onSuccess($params : Object; $models : cs.event.models)
        Function onData($request : 4D.HTTPRequest; $event : Object)
        Function onResponse($request : 4D.HTTPRequest; $event : Object)
        Function onTerminate($worker : 4D.SystemWorker; $params : Object)
    */
    
    $event.onError:=Formula(ALERT($2.message))
    $event.onSuccess:=Formula(ALERT($2.models.extract("name").join(",")+" loaded!"))
    $event.onData:=Formula(LOG EVENT(Into 4D debug message; This.file.fullName+":"+String((This.range.end/This.range.length)*100; "###.00%")))
    $event.onData:=Formula(MESSAGE(This.file.fullName+":"+String((This.range.end/This.range.length)*100; "###.00%")))
    $event.onResponse:=Formula(LOG EVENT(Into 4D debug message; This.file.fullName+":download complete"))
    $event.onResponse:=Formula(MESSAGE(This.file.fullName+":download complete"))
    $event.onTerminate:=Formula(LOG EVENT(Into 4D debug message; (["process"; $1.pid; "terminated!"].join(" "))))
    
    $port:=8085
    
    $options:={host: "127.0.0.1"}
    var $huggingfaces : cs.event.huggingfaces
    
    $folder:=$homeFolder.folder("Qwen3-4B-Thinking-2507")
    $path:="keisuke-miyako/Qwen3-4B-Thinking-2507-mlx-4bit"
    $URL:="keisuke-miyako/Qwen3-4B-Thinking-2507-mlx-4bit"
    
    $chat:=cs.event.huggingface.new($folder; $URL; $path)
    $huggingfaces:=cs.event.huggingfaces.new([$chat])
    
    $swama:=cs.swama.swama.new($port; $huggingfaces; $homeFolder; $options; $event)
    
End if 

Now you can test the server:

curl -X POST http://localhost:8080/v1/embeddings \
  -H "Content-Type: application/json" \
  -d '{
    "input": ["Hello world", "Text embeddings"],
    "model": "mlx-community/Qwen3-Embedding-0.6B-8bit"
  }'

curl -X POST http://localhost:8080/v1/chat/completions \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gemma3",
    "messages": [
      {"role": "user", "content": "Hello!"}
    ],
    "temperature": 0.7,
    "max_tokens": 100,
    "stream": true
  }'

Or, use AI Kit:

var $AIClient : cs.AIKit.OpenAI
$AIClient:=cs.AIKit.OpenAI.new()
$AIClient.baseURL:="http://127.0.0.1:8080/v1"

var $text : Text
$text:="The quick brown fox jumps over the lazy dog."

var $responseEmbeddings : cs.AIKit.OpenAIEmbeddingsResult
$responseEmbeddings:=$AIClient.embeddings.create($text)

Finally to terminate the server:

var $swama : cs.swama.swama
$swama:=cs.swama.swama.new()
$swama.terminate()

AI Kit compatibility

The API is compatibile with Open AI.

Class	API	Availability
Models	/v1/models	✅
Chat	/v1/chat/completions	✅
Images	/v1/images/generations
Moderations	/v1/moderations
Embeddings	/v1/embeddings	✅
Files	/v1/files