Classes
class	AILIAChatMessage

class	AILIALLMMediaData
	Media data structure for multimodal processing. More...

class	AILIALLMMultimodalChatMessage
	Multimodal chat message with media attachments. More...

Public Member Functions
static int	ailiaLLMCreate (ref IntPtr llm)
	Creates a LLM instance. More...

static int	ailiaLLMOpenModelFile (IntPtr llm, string path, uint n_ctx)
	Open model file. More...

static int	ailiaLLMOpenMtpModelFile (IntPtr llm, string path, uint n_draft)
	Open an MTP assistant and enable draft/target verification decoding. More...

static int	ailiaLLMSetSamplingParams (IntPtr llm, uint top_k, float top_p, float temp, uint dist)
	Set the sampling parameter. More...

static int	ailiaLLMSetThinking (IntPtr llm, uint enable)
	Enable or disable thinking (reasoning output). More...

static int	ailiaLLMSetTools (IntPtr llm, IntPtr tools_json)
	Set the tool (function) definitions. More...

static int	ailiaLLMSetPrompt (IntPtr llm, IntPtr messages, uint messages_len)
	Set the prompt. More...

static int	ailiaLLMSetPromptJson (IntPtr llm, IntPtr messages_json)

static int	ailiaLLMGetResponseJsonSize (IntPtr llm, ref uint size)

static int	ailiaLLMGetResponseJson (IntPtr llm, IntPtr json, uint size)

static int	ailiaLLMGenerate (IntPtr llm, ref uint done)
	Perform generate. More...

static int	ailiaLLMGetDeltaTextSize (IntPtr llm, ref uint len)
	Gets the size of text. (Include null) More...

static int	ailiaLLMGetDeltaText (IntPtr llm, IntPtr text, uint len)
	Gets the decoded text. More...

static int	ailiaLLMGetTokenCount (IntPtr llm, ref uint cnt, IntPtr text)
	Gets the count of token. More...

static int	ailiaLLMGetPromptTokenCount (IntPtr llm, ref uint cnt)
	Gets the count of prompt token. More...

static int	ailiaLLMGetGeneratedTokenCount (IntPtr llm, ref uint cnt)
	Gets the count of prompt token. More...

static int	ailiaLLMGetContextSize (IntPtr llm, ref uint context_size)
	Gets the size of context. More...

static int	ailiaLLMOpenMultimodalProjectorFile (IntPtr llm, string mmproj_path)
	Load multimodal projector file. More...

static int	ailiaLLMGetMultimodalCapabilities (IntPtr llm, ref uint vision_support, ref uint audio_support)
	Check if multimodal features are supported. More...

static int	ailiaLLMSetMultimodalPrompt (IntPtr llm, IntPtr messages, uint messages_len)
	Set multimodal prompt. More...

static int	ailiaLLMGetBackendCount (ref uint env_count)
	Gets the number of available computational environments (GPU, CPU, HTP (QNN)). More...

static int	ailiaLLMGetBackendName (ref IntPtr env, uint env_idx)
	Gets the list of computational environments. More...

static int	ailiaLLMGetBackendDeviceName (ref IntPtr name, uint env_idx)

static int	ailiaLLMSetBackend (IntPtr llm, uint backend_idx)

static int	ailiaLLMSetNumThreads (IntPtr llm, uint n_threads)
	Sets the CPU thread count before opening the model. More...

static void	ailiaLLMDestroy (IntPtr llm)
	It destroys the LLM instance. More...

static int	ailiaLLMGetQNNModelName (ref IntPtr model_name)
	Gets the device QNN artifact stem (for example, "sm8475") without a model. More...

static IntPtr	ailiaLLMGetErrorDetail (IntPtr llm)
	Gets the model's error detail as a library-owned UTF-8 string; do not free it. More...

Static Public Attributes
const String	LIBRARY_NAME ="ailia_llm"

const int	AILIA_LLM_STATUS_SUCCESS = (0)

const int	AILIA_LLM_STATUS_INVALID_ARGUMENT = (-1)

const int	AILIA_LLM_STATUS_ERROR_FILE_API = (-2)

const int	AILIA_LLM_STATUS_INVALID_VERSION = (-3)

const int	AILIA_LLM_STATUS_BROKEN = (-4)

const int	AILIA_LLM_STATUS_MEMORY_INSUFFICIENT = (-5)

const int	AILIA_LLM_STATUS_THREAD_ERROR = (-6)

const int	AILIA_LLM_STATUS_INVALID_STATE = (-7)

const int	AILIA_LLM_STATUS_CONTEXT_FULL = (-8)

const int	AILIA_LLM_STATUS_ERROR_BUFFER_API = (-9)

const int	AILIA_LLM_STATUS_PARSE_ERROR = (-10)

const int	AILIA_LLM_STATUS_UNIMPLEMENTED = (-15)

const int	AILIA_LLM_STATUS_OTHER_ERROR = (-128)

Member Function Documentation

◆ ailiaLLMCreate()

static int ailiaLLM.AiliaLLM.ailiaLLMCreate ( ref IntPtr llm )

Creates a LLM instance.

Parameters

llm	A pointer to the LLM instance pointer

Returns: If this function is successful, it returns AILIA_LLM_STATUS_SUCCESS , or an error code otherwise.

Creates a LLM instance.

Remarks: When using QNN, the LLM instance must be destroyed by calling ailiaLLMDestroy on the same thread that called this function.

◆ ailiaLLMDestroy()

static void ailiaLLM.AiliaLLM.ailiaLLMDestroy ( IntPtr llm )

It destroys the LLM instance.

Parameters

llm	A LLM instance pointer

Remarks: When using QNN, this function must be called on the same thread that called ailiaLLMCreate. When using QNN, moving only the release operation to another thread, such as the UI thread, may cause QnnDevice_create failed: code 14001 when QNN is subsequently used on a different thread.

◆ ailiaLLMGenerate()

static int ailiaLLM.AiliaLLM.ailiaLLMGenerate	(	IntPtr	llm,
		ref uint	done
	)

Perform generate.

Parameters

llm	A LLM instance pointer
done	Generation complete?

Returns: If this function is successful, it returns AILIA_LLM_STATUS_SUCCESS , or an error code otherwise.

The decoded result is obtained through the ailiaLLMGetDeltaText API. Each call to ailiaLLMGenerate decodes one token at a time. The value of done is 0 or 1. If done is 1, the generation is complete.

◆ ailiaLLMGetBackendCount()

static int ailiaLLM.AiliaLLM.ailiaLLMGetBackendCount ( ref uint env_count )

Gets the number of available computational environments (GPU, CPU, HTP (QNN)).

Parameters

env_count The storage location of the number of computational environment information

Returns: If this function is successful, it returns AILIA_LLM_STATUS_SUCCESS , or an error code otherwise.

◆ ailiaLLMGetBackendDeviceName()

static int ailiaLLM.AiliaLLM.ailiaLLMGetBackendDeviceName	(	ref IntPtr	name,
		uint	env_idx
	)

◆ ailiaLLMGetBackendName()

static int ailiaLLM.AiliaLLM.ailiaLLMGetBackendName	(	ref IntPtr	env,
		uint	env_idx
	)

Gets the list of computational environments.

Parameters

env	The storage location of the computational environment information (valid until the AILIANetwork instance is destroyed)
env_idx	The index of the computational environment information (0 to ailiaLLMGetBackendCount() -1)

Returns: If this function is successful, it returns AILIA_LLM_STATUS_SUCCESS , or an error code otherwise.

◆ ailiaLLMGetContextSize()

static int ailiaLLM.AiliaLLM.ailiaLLMGetContextSize	(	IntPtr	llm,
		ref uint	context_size
	)

Gets the size of context.

Parameters

llm	A LLM instance pointer
context_size	The length of context

Returns: If this function is successful, it returns AILIA_LLM_STATUS_SUCCESS , or an error code otherwise.

◆ ailiaLLMGetDeltaText()

static int ailiaLLM.AiliaLLM.ailiaLLMGetDeltaText	(	IntPtr	llm,
		IntPtr	text,
		uint	len
	)

Gets the decoded text.

Parameters

llm	A LLM instance pointer
text	Text(UTF8)
len	Buffer size

Returns: If this function is successful, it returns AILIA_STATUS_SUCCESS , or an error code otherwise.

If ailiaLLMGenerate() is not run at all, the function returns AILIA_STATUS_INVALID_STATE .

◆ ailiaLLMGetDeltaTextSize()

static int ailiaLLM.AiliaLLM.ailiaLLMGetDeltaTextSize	(	IntPtr	llm,
		ref uint	len
	)

Gets the size of text. (Include null)

Parameters

llm	A LLM instance pointer
len	The length of text

Returns: If this function is successful, it returns AILIA_STATUS_SUCCESS , or an error code otherwise.

◆ ailiaLLMGetErrorDetail()

static IntPtr ailiaLLM.AiliaLLM.ailiaLLMGetErrorDetail ( IntPtr llm )

Gets the model's error detail as a library-owned UTF-8 string; do not free it.

Parameters

llm	An LLM instance pointer

Returns: Pointer to a library-owned UTF-8 string

◆ ailiaLLMGetGeneratedTokenCount()

static int ailiaLLM.AiliaLLM.ailiaLLMGetGeneratedTokenCount	(	IntPtr	llm,
		ref uint	cnt
	)

Gets the count of prompt token.

Parameters

llm	A LLM instance pointer
cnt	The count of generated token

Returns: If this function is successful, it returns AILIA_LLM_STATUS_SUCCESS , or an error code otherwise.

It can be called after calling ailiaLLMGenerate.

◆ ailiaLLMGetMultimodalCapabilities()

static int ailiaLLM.AiliaLLM.ailiaLLMGetMultimodalCapabilities	(	IntPtr	llm,
		ref uint	vision_support,
		ref uint	audio_support
	)

Check if multimodal features are supported.

Parameters

llm	A LLM instance pointer
vision_support	Whether image input (VLM) is supported
audio_support	Whether audio input (ALM) is supported

Returns: If this function is successful, it returns AILIA_LLM_STATUS_SUCCESS , or an error code otherwise.

Can be called after ailiaLLMOpenMultimodalProjectorFile.

◆ ailiaLLMGetPromptTokenCount()

static int ailiaLLM.AiliaLLM.ailiaLLMGetPromptTokenCount	(	IntPtr	llm,
		ref uint	cnt
	)

Gets the count of prompt token.

Parameters

llm	A LLM instance pointer
cnt	The count of prompt token

Returns: If this function is successful, it returns AILIA_LLM_STATUS_SUCCESS , or an error code otherwise.

It can be called after calling ailiaLLMSetPrompt.

◆ ailiaLLMGetQNNModelName()

static int ailiaLLM.AiliaLLM.ailiaLLMGetQNNModelName ( ref IntPtr model_name )

Gets the device QNN artifact stem (for example, "sm8475") without a model.

Parameters

model_name Pointer to a library-owned string; do not free it.

Returns: UNIMPLEMENTED when QNN is unavailable, or OTHER_ERROR for unsupported devices.

QNN is supported on Android arm64 and Windows ARM64.

◆ ailiaLLMGetResponseJson()

static int ailiaLLM.AiliaLLM.ailiaLLMGetResponseJson	(	IntPtr	llm,
		IntPtr	json,
		uint	size
	)

◆ ailiaLLMGetResponseJsonSize()

static int ailiaLLM.AiliaLLM.ailiaLLMGetResponseJsonSize	(	IntPtr	llm,
		ref uint	size
	)

◆ ailiaLLMGetTokenCount()

static int ailiaLLM.AiliaLLM.ailiaLLMGetTokenCount	(	IntPtr	llm,
		ref uint	cnt,
		IntPtr	text
	)

Gets the count of token.

Parameters

llm	A LLM instance pointer
cnt	The count of token
text	Text(UTF8)

Returns: If this function is successful, it returns AILIA_LLM_STATUS_SUCCESS , or an error code otherwise.

◆ ailiaLLMOpenModelFile()

static int ailiaLLM.AiliaLLM.ailiaLLMOpenModelFile	(	IntPtr	llm,
		string	path,
		uint	n_ctx
	)

Open model file.

Parameters

llm	A pointer to the LLM instance pointer
path	Path for a GGUF model or QNN package
n_ctx	Context length for model (0 is model default）

Returns: If this function is successful, it returns AILIA_STATUS_SUCCESS , or an error code otherwise.

Without an explicit backend selection, GGUF selects CPU/GPU and .qnn selects HTP. A model format mismatching an explicit selection is rejected.

◆ ailiaLLMOpenMtpModelFile()

static int ailiaLLM.AiliaLLM.ailiaLLMOpenMtpModelFile	(	IntPtr	llm,
		string	path,
		uint	n_draft
	)

Open an MTP assistant and enable draft/target verification decoding.

Parameters

llm	LLM instance
path	Path to a matching Gemma 4 assistant GGUF, or assistant .qnn for a QNN target
n_draft	Number of draft tokens (at least 1). QNN supports up to 3.

Call once after opening the target and before setting a prompt. Supports text generation with CPU/GPU Gemma 4 GGUF and QNN models. Verify the requested draft tokens together with one target token.

◆ ailiaLLMOpenMultimodalProjectorFile()

static int ailiaLLM.AiliaLLM.ailiaLLMOpenMultimodalProjectorFile	(	IntPtr	llm,
		string	mmproj_path
	)

Load multimodal projector file.

Parameters

llm	A LLM instance pointer
mmproj_path	Path to an MMPROJ GGUF or self-contained ailia QNN projector (.qnn)

Returns: If this function is successful, it returns AILIA_LLM_STATUS_SUCCESS , or an error code otherwise.

To use multimodal features, you must first load the text model with ailiaLLMOpenModelFile, then load the multimodal projector with this function. Image input (VLM) requires a vision-capable projector, while audio input (ALM) requires an audio-capable projector. QNN projectors support Android arm64 and Windows ARM64.

◆ ailiaLLMSetBackend()

static int ailiaLLM.AiliaLLM.ailiaLLMSetBackend	(	IntPtr	llm,
		uint	backend_idx
	)

◆ ailiaLLMSetMultimodalPrompt()

static int ailiaLLM.AiliaLLM.ailiaLLMSetMultimodalPrompt	(	IntPtr	llm,
		IntPtr	messages,
		uint	messages_len
	)

Set multimodal prompt.

Parameters

llm	A LLM instance pointer
message	Array of multimodal messages
message_cnt	Number of messages

Returns: If this function is successful, it returns AILIA_LLM_STATUS_SUCCESS , or an error code otherwise.

Set multimodal prompt. Include <media> placeholders in message content, and set corresponding media data in media_data. Example: "Describe this image: <__media__>"

◆ ailiaLLMSetNumThreads()

static int ailiaLLM.AiliaLLM.ailiaLLMSetNumThreads	(	IntPtr	llm,
		uint	n_threads
	)

Sets the CPU thread count before opening the model.

Parameters

llm	LLM instance
n_threads	Thread count (0 selects automatically).

Returns: AILIA_LLM_STATUS_SUCCESS on success, or AILIA_LLM_STATUS_INVALID_STATE after a model is opened.

The default is 0 when this API is not called. On Android, automatic selection uses the number of performance (big/prime) cores. If topology is unavailable, the existing core-count heuristic is used. Target decode/prefill, assistant and VLM/ALM CPU inference threads share this value.

◆ ailiaLLMSetPrompt()

static int ailiaLLM.AiliaLLM.ailiaLLMSetPrompt	(	IntPtr	llm,
		IntPtr	messages,
		uint	messages_len
	)

Set the prompt.

Parameters

llm	A pointer to the LLM instance pointer
message	Array of messages
message_cnt	Number of messages

Returns: If this function is successful, it returns AILIA_STATUS_SUCCESS , or an error code otherwise.

Set the prompt to query the LLM. Please include ChatHistory in the message as well.

◆ ailiaLLMSetPromptJson()

static int ailiaLLM.AiliaLLM.ailiaLLMSetPromptJson	(	IntPtr	llm,
		IntPtr	messages_json
	)

◆ ailiaLLMSetSamplingParams()

static int ailiaLLM.AiliaLLM.ailiaLLMSetSamplingParams	(	IntPtr	llm,
		uint	top_k,
		float	top_p,
		float	temp,
		uint	dist
	)

Set the sampling parameter.

Parameters

llm	A pointer to the LLM instance pointer
top_k	Sampling probability value's top number, default 40
top_p	Sampling probability value range, default 0.9 (0.9 to 1.0)
temp	Temperature parameter, default 0.4
dist	Seed, default 1234

Returns: If this function is successful, it returns AILIA_LLM_STATUS_SUCCESS , or an error code otherwise.

Set LLM sampling parameters. Must be run before ailiaLLMSetPrompt.

◆ ailiaLLMSetThinking()

static int ailiaLLM.AiliaLLM.ailiaLLMSetThinking	(	IntPtr	llm,
		uint	enable
	)

Enable or disable thinking (reasoning output).

Parameters

llm	A pointer to the LLM instance pointer
enable	0 to disable, non-zero to enable (default: disabled)

Returns: If this function is successful, it returns AILIA_LLM_STATUS_SUCCESS , or an error code otherwise.

Controls whether thinking models (e.g. Gemma4) output their reasoning process. Must be called before ailiaLLMSetPrompt.

◆ ailiaLLMSetTools()

static int ailiaLLM.AiliaLLM.ailiaLLMSetTools	(	IntPtr	llm,
		IntPtr	tools_json
	)

Set the tool (function) definitions.

Parameters

llm	A LLM instance pointer
tools_json	OpenAI-compatible JSON array of tool definitions (UTF-8, null terminated). IntPtr.Zero clears the tools.

Returns: If this function is successful, it returns AILIA_LLM_STATUS_SUCCESS , or an error code otherwise.

Tools are defined in the same format as the tools parameter of the OpenAI Chat Completions API. They are rendered into the prompt through the chat template on the next ailiaLLMSetPrompt, and the output is constrained by a grammar for the tool call syntax. Parse the raw output with ailiaLLMGetResponseJson. While tools are set, the content of role "assistant" is the raw output and the content of role "tool" is the tool result. Legacy prompt APIs reject tool use. Use SetPromptJson/GetResponseJson instead.

Member Data Documentation

◆ AILIA_LLM_STATUS_BROKEN

const int ailiaLLM.AiliaLLM.AILIA_LLM_STATUS_BROKEN = (-4)

static

◆ AILIA_LLM_STATUS_CONTEXT_FULL

const int ailiaLLM.AiliaLLM.AILIA_LLM_STATUS_CONTEXT_FULL = (-8)

static

◆ AILIA_LLM_STATUS_ERROR_BUFFER_API

const int ailiaLLM.AiliaLLM.AILIA_LLM_STATUS_ERROR_BUFFER_API = (-9)

static

◆ AILIA_LLM_STATUS_ERROR_FILE_API

const int ailiaLLM.AiliaLLM.AILIA_LLM_STATUS_ERROR_FILE_API = (-2)

static

◆ AILIA_LLM_STATUS_INVALID_ARGUMENT

const int ailiaLLM.AiliaLLM.AILIA_LLM_STATUS_INVALID_ARGUMENT = (-1)

static

◆ AILIA_LLM_STATUS_INVALID_STATE

const int ailiaLLM.AiliaLLM.AILIA_LLM_STATUS_INVALID_STATE = (-7)

static

◆ AILIA_LLM_STATUS_INVALID_VERSION

const int ailiaLLM.AiliaLLM.AILIA_LLM_STATUS_INVALID_VERSION = (-3)

static

◆ AILIA_LLM_STATUS_MEMORY_INSUFFICIENT

const int ailiaLLM.AiliaLLM.AILIA_LLM_STATUS_MEMORY_INSUFFICIENT = (-5)

static

◆ AILIA_LLM_STATUS_OTHER_ERROR

const int ailiaLLM.AiliaLLM.AILIA_LLM_STATUS_OTHER_ERROR = (-128)

static

◆ AILIA_LLM_STATUS_PARSE_ERROR

const int ailiaLLM.AiliaLLM.AILIA_LLM_STATUS_PARSE_ERROR = (-10)

static

◆ AILIA_LLM_STATUS_SUCCESS

const int ailiaLLM.AiliaLLM.AILIA_LLM_STATUS_SUCCESS = (0)

static

◆ AILIA_LLM_STATUS_THREAD_ERROR

const int ailiaLLM.AiliaLLM.AILIA_LLM_STATUS_THREAD_ERROR = (-6)

static

◆ AILIA_LLM_STATUS_UNIMPLEMENTED

const int ailiaLLM.AiliaLLM.AILIA_LLM_STATUS_UNIMPLEMENTED = (-15)

static

◆ LIBRARY_NAME

const String ailiaLLM.AiliaLLM.LIBRARY_NAME ="ailia_llm"

static

The documentation for this class was generated from the following file:

AiliaLLM.cs

Classes

Public Member Functions

Static Public Attributes

Member Function Documentation

◆ ailiaLLMCreate()

◆ ailiaLLMDestroy()

◆ ailiaLLMGenerate()

◆ ailiaLLMGetBackendCount()

◆ ailiaLLMGetBackendDeviceName()

◆ ailiaLLMGetBackendName()

◆ ailiaLLMGetContextSize()

◆ ailiaLLMGetDeltaText()

◆ ailiaLLMGetDeltaTextSize()

◆ ailiaLLMGetErrorDetail()

◆ ailiaLLMGetGeneratedTokenCount()

◆ ailiaLLMGetMultimodalCapabilities()

◆ ailiaLLMGetPromptTokenCount()

◆ ailiaLLMGetQNNModelName()

◆ ailiaLLMGetResponseJson()

◆ ailiaLLMGetResponseJsonSize()

◆ ailiaLLMGetTokenCount()

◆ ailiaLLMOpenModelFile()

◆ ailiaLLMOpenMtpModelFile()

◆ ailiaLLMOpenMultimodalProjectorFile()

◆ ailiaLLMSetBackend()

◆ ailiaLLMSetMultimodalPrompt()

◆ ailiaLLMSetNumThreads()

◆ ailiaLLMSetPrompt()

◆ ailiaLLMSetPromptJson()

◆ ailiaLLMSetSamplingParams()

◆ ailiaLLMSetThinking()

◆ ailiaLLMSetTools()

Member Data Documentation

◆ AILIA_LLM_STATUS_BROKEN

◆ AILIA_LLM_STATUS_CONTEXT_FULL

◆ AILIA_LLM_STATUS_ERROR_BUFFER_API

◆ AILIA_LLM_STATUS_ERROR_FILE_API

◆ AILIA_LLM_STATUS_INVALID_ARGUMENT

◆ AILIA_LLM_STATUS_INVALID_STATE

◆ AILIA_LLM_STATUS_INVALID_VERSION

◆ AILIA_LLM_STATUS_MEMORY_INSUFFICIENT

◆ AILIA_LLM_STATUS_OTHER_ERROR

◆ AILIA_LLM_STATUS_PARSE_ERROR

◆ AILIA_LLM_STATUS_SUCCESS

◆ AILIA_LLM_STATUS_THREAD_ERROR

◆ AILIA_LLM_STATUS_UNIMPLEMENTED

◆ LIBRARY_NAME