LlamaMBS.ApplyChatTemplate(Template as String, ChatMessages() as LlamaChatMessageMBS, AddAssistentTokens as boolean) as String

Template: The Jinja template to use for this chat.
ChatMessages: a list of multiple llama_chat_message
AddAssistentTokens: Whether to end the prompt with the token(s) that indicate the start of an assistant message.

Returns the answer. May raise LlamaExceptionMBS class if the apply process fails.

Some examples using this method:

• /Tools/Llama/LLama Simple Chat

LlamaMBS.BackendInit

If numa is true, use NUMA optimizations
Call once at the start of the program

LlamaMBS.BackendLoad(Path as String) as Boolean

Please pass DLL, dylib or so path.

LlamaMBS.BackendLoadAll

// load dynamic backends

LlamaMBS.BackendLoadAll

Needs ggml library loaded.

Some examples using this method:

• /Tools/Llama/LLama Simple
• /Tools/Llama/LLama Simple Chat
• /Tools/Llama/LLama Simple Chat Gemma 4
• /Tools/Llama/LLama Simple console

LlamaMBS.Backends as LlamaBackendMBS()

// load dynamic backends
LlamaMBS.BackendLoadAll

// now inspect them
var backends() as LlamaBackendMBS = LlamaMBS.Backends
for each backend as LlamaBackendMBS in backends
  var devices() as LlamaDeviceMBS = backend.Devices
  var name as string = backend.Name
  var features as Dictionary = backend.Features
  
  Break
next

Each backend has a list of devices with properties.

Some examples using this method:

• /Tools/Llama/LLama Simple
• /Tools/Llama/LLama Simple Chat
• /Tools/Llama/LLama Simple Chat Gemma 4
• /Tools/Llama/LLama Simple console

LlamaMBS.ChatBuiltInTemplates as String()

LlamaMBS.Loaded as Boolean

LlamaMBS.LoadErrorMessage as String

LlamaMBS.LoadLibrary(File as FolderItem) as boolean

For Windows you need llama.dll, for macOS you need llama.dylib and for Linux you need llama.so.

On Windows you need to load ggml.dll, ggml-base.dll and llama.dll to have all functions.

Returns true on success and false on failure.

See also:

• LoadLibrary(Path as String) as boolean

LlamaMBS.LoadLibrary(Path as String) as boolean

If Not LlamaMBS.LoadLibrary("/opt/homebrew/Cellar/llama.cpp/6710/lib/libllama.dylib") Then
  System.DebugLog LlamaMBS.LoadErrorMessage
End If

For Windows you need llama.dll, for macOS you need llama.dylib and for Linux you need llama.so.

On Windows you need to load ggml.dll, ggml-base.dll and llama.dll to have all functions.

Returns true on success and false on failure.

See also:

• LoadLibrary(File as FolderItem) as boolean

LlamaMBS.PrintSystemInfo

LlamaMBS.SetDebugLogger

Helps to debug console applications.

LlamaMBS.SetLogger(Logger as LlamaLogMBS)

If no logger is set, log messages go to stderr.

Some examples using this method:

• /Tools/Llama/LLama Simple console