API ReferencesDiscordGlitch9
Edit

AI Dev Kit v3

v.3.11.22 (Jun 26, 2025) – Preview

  • Added PixelLab API (Pro-only preview; may move to addon)

  • Fixed streaming LLM response issues

  • Upgraded multiple Generator UIs

v.3.11.20 (Jun 20, 2025) – Preview

  • Refactored EndpointType(int) → RequestType(string)

  • Added TTS + Transcript streaming

  • Added new LogProb class

  • Renamed RealtimeAudioPlayer → StreamingAudioPlayer

  • Extended StreamingAudioPlayer (input sample rate, crackling remover, UnityEvents)

  • Added TTS Streaming Demo + Demo 8

v.3.11.18–19 (Jun 19, 2025) – Hotfix

  • Potential fixes for IL2CPP build issues

v.3.11.17 (May 29, 2025) – Hotfix

  • Fixed Speech Generator stack overflow

  • Fixed AddressableUtils issues

  • Added Google Voices + TTS Tasks

v.3.11.16 (May 29, 2025) – Preview

  • New Generator Windows tested/debugged

  • Added multiple outputs for GeneratedText + Moderation

  • Added GridView select-action

  • Abstracted GeneratorWindowSettings

  • Added multi-selection in Generator Windows

  • Removed obsolete GENImageVariationTask

v.3.11.15 (May 28, 2025) – Preview

  • New Generators: Avatar, Background, Code, SFX, Video, Component

  • Added yield/await to all GENTasks

  • Added extra outputs to prompt records

  • Fixed Prompt Field resize handle

  • Updated base GeneratorWindow class

v.3.11.14 (May 27, 2025) – Preview

  • New EditorTool layout test

v.3.11.13 (May 24, 2025) – Preview

  • New Modular Chatbot component

    • Integrates STT (input), TTS (output), ImageGen

    • Auto-resolves response type (TTS/Image)

    • Internal AudioSource for TTS playback

v.3.11.12 (May 24, 2025) – Preview

  • New Components: TTS, STT, Image Generator

  • Fixed GoogleFile upload + MIMEType issues

  • Fixed some icon path issues

v.3.11.11 (May 24, 2025) – Preview

  • Fixed EditorChat + GENChatTask Unity crash

v.3.11.10 (May 24, 2025) – Preview

⚠️ Major refactor (clean install required)

  • IUniFile → IFile

  • UniFileBase<T> → File<T>

  • UniAudioFile → File<AudioClip>

  • UniImageFile → File<Texture2D>

  • UniVideoFile, UniFile → RawFile

  • AIProvider → Api

  • UploadedFile → ApiFile

  • RealtimeAudioStreamer → RealtimeAudioPlayer

  • Removed FormFile, SerializedContent

v.3.11.09 (May 22, 2025) – Hotfix

  • Fixed Generator Window not saving model change

v.3.11.08 (May 22, 2025) – Hotfix

  • Fixed Ollama streaming issues

v.3.11.07 (May 21, 2025) – Preview

  • Assistants API integration v3 (tested)

  • Realtime API integration v2 (tested)

  • Added WebSocketEventReceiver, ToolEventReceiver, StreamingAudioEventReceiver

  • Added Realtime + Receiver icons

  • Moved GENChat + ChatSession to Pro-only

  • Fixed catalogue duplication issues

  • Upgraded model window UI

  • Fixed demo scenes

v.3.11.06 (May 21, 2025) – Preview

  • Split Chat Event Receiver into two classes:

    • Chat Event Receiver (string)

    • Chat Event Receiver (ChatMessage)

v.3.11.05 (May 21, 2025) – Preview

  • Added save folder button in Chatbot inspector

  • Added TTS to SpeechAPI

  • Fixed expandable text field

v.3.11.04 (May 21, 2025) – Preview

  • Multiple major changes:

    • FunctionManager supports return values

    • Google File Manager implemented (upload + management)

    • Added VoiceLibrary filter window

    • Added multi-response support

    • Forced model/voice dropdown update

  • Renamed:

    • ModelCapability → ModelFeature

    • AIDevKit.FileData → UploadedFile

    • Google.FileData → GoogleFileData

    • Google.File → GoogleFile

    • OpenAI.File → OpenAIFile

  • Removed old/deprecated classes (e.g., UploadFileRequest)

v.3.11.03 (May 21, 2025) – Hotfix

  • Fixed duplicate entry issue in models window

  • Renamed tasks:

    • GENImageEdit → GENInpaint

    • GENTranscript now returns Transcript

    • GENTranslation now returns string

v.3.11.02 (May 21, 2025) – Preview

  • Introduced GENSequence prototype

  • Added GENModerationTask + Image Moderation (OpenAI)

  • Implemented UnityEngine UI extension methods for GENTask

  • Removed Google GenerativeAIModel + ChatSession

  • Cleaned up CRUDClient/Service

v.3.11.01 (May 21, 2025) – Preview

  • Added ChatSession.Rewind() (Google-style)

  • Added moderation handling improvements

  • Merged GeneratedContent → ChatCompletion

v.3.11.00 (May 21, 2025) – Preview

  • Major refactor of CRUDClient + CRUDService

  • Refactored ChatMessage class into: UserMessage, SystemMessage, AssistantMessage, ToolMessage

  • Integrated moderation into GENChat + Chatbot (new ModerationHandler)

  • Updated demo scenes

v.3.10.04 (Jun 5, 2025) – Preview

  • Added entry deletion to Model Catalogue

  • Added “Delete All Deprecated Models” option

  • Added Custom Model deletion support

  • Special row text colors for custom/obsolete models

  • Added Chatbot inspector redesign for beginners

  • Added Moderator module to Chatbot

  • Minor bug fixes

v.3.10.01 (Jun 4, 2025) – Preview

  • Fixed conversion issues

  • Added “Edit Script with AI” (Inspector right-click)

  • Added silence detection to Speech To Text component

v.3.8.21 (Jun 1, 2025) – Preview

  • Fixed build issues

  • Cleaned editor scripts

v.3.8.20 (May 30, 2025) – Preview

  • Fixed Ollama server connection issue

  • Fixed model update window termination issue

v.3.8.19 (May 30, 2025) – Preview

  • Added click-to-show prompt history details in Generator Window

  • Fixed total price calculation in Prompt History window

v.3.8.18 (May 30, 2025) – Preview

  • Tested & debugged Google TTS

  • GeneratedAsset → GeneratedFile

  • Generator Windows remember last prompt state

  • Fixed Unity Component Generator

  • Fixed Code Generator

  • Upgraded Generator Windows UI

  • Fixed Google MIMEType parsing issue

  • Changed STT component Transcript receiver → Text receiver (Transcript still available via code)

v.3.8.17 (May 29, 2025) – Preview

  • Hotfix for Speech Generator stack overflow

  • Fixed AddressableUtils issues

  • Added Google Voices

  • Added Google TTS Tasks

v.3.8.16 (May 29, 2025) – Preview

  • All new Generator Windows tested & debugged

  • GeneratedText: string, string[], usage

  • Moderation: SafetySetting, SafetySetting[], usage

  • Fixed video prompt record showing as TTS record

  • Added select-action to GridView

  • Abstracted GeneratorWindowSettings

  • Added multi-selection to Generator Windows

  • Merged AIDevKitSettings.RuntimeOutputPath + EditorOutputPath → OutputPath

  • Removed GENImageVariationTask (DALL-E2 only, deprecated)

v.3.8.15 (May 28, 2025) – Preview

  • Added Avatar Generator

  • Added Background Generator

  • Added yield await to all GENTasks

  • Added method to append extra outputs to prompt records

  • Added GENCode (generate C# code for Unity)

  • Added new Generator GUIs: Code, SFX, Video, Component

  • Fixed Prompt Field resize handle

  • Added GENVideo prompt records

  • BaseGeneratorWindow → GeneratorWindowBase

  • Added “Animated Toggle IMGUI”

v.3.8.14 (May 27, 2025) – Preview

  • New EditorTool layout test

v.3.8.13 (May 24, 2025) – Preview

  • Added Modular Chatbot (inherits from Chatbot)

    • Supports Speech-to-Text for voice input

    • Supports Text-to-Speech for voice output

    • Supports Image generation in responses

    • Auto-resolves response type and plays TTS or renders images

    • Internal AudioSource handling for TTS playback

v.3.8.12 (May 24, 2025) – Preview

  • Added new Components: Text-to-Speech, Speech-to-Text, Image Generator

  • Fixed GoogleFile upload issues

  • Fixed GoogleFile MIMEType issues

  • Fixed some icon path issues

v.3.8.11 (May 24, 2025) – Preview

  • Fixed EditorChat issues

  • Fixed GENChatTask Unity crash

v.3.8.09 (May 24, 2025) – Preview

⚠️ Clean install required due to core refactoring

  • IUniFile → IFile

  • UniFileBase<T> → File<T> (abstract → class)

  • UniAudioFile → File<AudioClip>

  • UniImageFile → File<Texture2D>

  • UniVideoFile → RawFile

  • UniFile → RawFile

  • AIProvider → Api

  • UploadedFile → ApiFile

  • RealtimeAudioStreamer → RealtimeAudioPlayer

  • StreamAudioPlayer → StreamingAudioPlayer

  • Removed FormFile, SerializedContent (replaced with RawFile, File<T>)

v.3.7.9 (May 21, 2025) – Preview

  • Split Chat Event Receiver into two derived classes:

    • Chat Event Receiver (string)

    • Chat Event Receiver (ChatMessage)

v.3.7.8 (May 21, 2025) – Preview

  • Added “open save folder” button in Chatbot inspector

  • Added TTS feature to SpeechAPI

  • Fixed expandable text field

v.3.7.7 (May 21, 2025) – Preview

  • Integrated Assistants API (v3) and Realtime API (v2) – tested

  • Added WebSocketEventReceiver, ToolEventReceiver, StreamingAudioEventReceiver

  • Added new icons for Realtime + Receivers

  • Moved GENChat, ChatSession to Pro-only

  • Fully resolved catalogue duplication issues

  • New/Deprecated Model window UI upgrade

  • Fixed demo scenes (except Moderation)

v.3.7.6 (May 29, 2025) – Preview

  • Upgraded Assistants API (v3) + Realtime API (v2)

  • Added WebSocketEventReceiver, ToolEventReceiver, StreamingAudioEventReceiver

  • Added new icons for Realtime + Receivers

  • Moved GENChat, ChatSession to Pro-only

  • Fully resolved catalogue duplication issues

  • New Model window UI; fixed demo scenes

v.3.7.5 (May 22, 2025) – Hotfix

  • Finished 80% of GitBook docs

  • Tested/re-ordered Generate menu for Unity6

  • EditorChat input auto-resizes

  • Added audio input + multi-attachments

  • Flattened RealtimeEventReceiver hierarchy

v.3.7.3 (May 21, 2025) – Hotfix

  • Fixed ChatRole conversion issue

  • Fixed ChatStreamHandler

v.3.7.2 (May 21, 2025) – Preview

  • Fixed duplicate model entries in window

  • GENImageEdit → GENInpaint

  • GENTranscript returns Transcript

  • GENTranslation returns string

v.3.7.1 (May 21, 2025) – Preview

  • Added ChatSession.Rewind()

  • Added GENModerationTask, Image Moderation

  • Added GENSequence prototype + Unity Object extension methods

  • Applied Google-style moderation handling

  • Cleaned up CRUD code

  • Removed Google GenerativeAI model impl + ChatSession

v.3.7.0 (May 21, 2025) – Preview

  • Added OpenAI Projects API

  • Integrated moderation into GENChat + Chatbot

  • Added UnityEvent-based receivers (NonStreamingChat, StreamingChat, Speech, Realtime)

  • Refactored CRUDClient/Service

  • Refactored ChatMessage into UserMessage/SystemMessage/AssistantMessage/ToolMessage

v.3.6.18–v.3.6.23 (May 18–23, 2025) – Hotfixes

  • Added WebSearchOptions, SpeechOutputOptions

  • Moved catalogue system to common assembly (Free compatible)

  • Changed catalogue update behavior (replace → add)

  • Improved key encryption + library GUIs

  • Improved Chatbot

  • Abstracted catalogues

  • Multiple hotfixes: Ollama server, Model Catalogue flags, EditorChat

  • Chatbot gained UnityEvents

v.3.6.14–v.3.6.16 (May 14–16, 2025) – Hotfixes

  • DebugMode for APIs + AIDevKit

  • Added STT Demo, ElevenLabs fixes

  • Multipart form fixes for ElevenLabs/Google

  • Realtime API moved to Pro-only, renamed to RealtimeAssistant

  • SpeechToTextAPI → SpeechAssistant (now STT/translation/voice changer)

  • ChatAPI → Chatbot

  • Added voice changer demo

  • Removed legacy components (ImageGenAPI, TTSAPI, GeminiAPI)

v.3.6.8–v.3.6.12 (May 8–12, 2025) – Stable hotfixes

  • Model Catalogue, Gemini, TTS, and LLM non-streaming fixes

v.3.6.7 (May 7, 2025) – Stable v7

  • Added reasoning to GENTask

  • Added toolcalls to GeneratedText/GeneratedContent

  • Added UniVideoFile

  • Added streaming usage return to all APIs

  • Fixed history saving, treeview updates, request timeouts

  • Improved EditorChat error handling

v.3.6.6 (May 6, 2025) – Stable v6

  • New ChatStreamHandler supporting tools/functions

  • Usage tracking for GENText, GENChat

  • Prompt History supports images & audio

  • Added GENVideo task + Editor menu

  • Merged Completion + ChatCompletion

  • Fixed Imagen, STT, and other API bugs

  • Refactored prompt history creation

v.3.6.4 (May 2, 2025) – Stable v3

  • Removed model price simulator

  • Merged Model Library + Catalogue

  • Merged Voice Library + Catalogue

  • Combined Model + ModelProfile, Voice + VoiceProfile

  • Added OpenAI metadata for models/voices

  • Auto-updated ModelSelector cache on additions

  • Fixed Editor Chat UX issues (input focus, flickering, autosave debounce)

v.3.6.3 (May 1, 2025) – Stable v2

  • Stabilized Model Selector

  • Added famous/stable models as default

  • Added cost info in Model Catalogue (OpenRouter only)

v.3.6.2 (May 1, 2025) – Stable

  • Fixed Assistants API & demo scenes

  • Added fallbacks for missing model entries

  • Improved model name analyzer

  • Enforced strict defaults for paths, models, and voices

  • Stabilization works

v.3.6.0 (Apr 28–29, 2025) – Unstable

  • Removed VoiceSelectorGUI; added static VoicePopupGUI

  • Improved meta resolvers

  • Added image + function support to GENChat

  • Added Ollama/OpenRouter integration with model families

  • Improved ModelSelectorGUI with defaults, flags, loading screens

v.3.5.3 (Apr 27, 2025) – Unstable

  • Redesigned & abstracted chat system

  • Removed StringOr<MessageContent> converters

  • Added custom path option to ScriptableDatabase

  • Remade chat demo scene

v.3.5.2 (Apr 24, 2025) – Unstable

  • Refactored core REST client for flexible API abstraction

  • Removed Query() in favor of List()

  • Refactored RESTRequest with body support

  • Added Ollama Embeddings, OpenRouter API, and more model management features

  • Unified and renamed multiple core classes (e.g., ToolDefinition → ToolCall)

v.3.5.1 (Apr 23, 2025)

  • Refined REST download / realtime load asset conditions

  • Wrote and added WebGL MPEG decoder

  • Fixed “Generate” menu output path

  • Refactored and redesigned EditorSpeech

  • Persisted GenWindows output path across compiles

  • Added troubleshooting docs for plugin update errors

v.3.5.0 (Apr 22, 2025)

  • Fixed vision prompt word-wrap and missing field issues

  • Enabled multi-image generation for models without native support

  • Eliminated GUI errors in EditorTools

  • Added drag & drop to editor vision

  • Fixed EditorSpeech result display

  • Redesigned EditorSpeech; output path now persists in GenWindows

  • Refined REST download / realtime asset load

  • Added WebGL MPEG decoder

  • Fixed “Generate” menu output path

  • Added troubleshooting docs for plugin update errors

v.3.4.19 (Apr 21, 2025)

  • Added ElevenLabs prompt history builder

  • Fixed CustomEditor not saving values

  • Integrated FunctionManager into ChatCompletionTool and ChatSessionTool

  • Fixed broken demo scenes

  • Fixed OpenAI prompt history builder

  • Fixed Google prompt history builder

  • Refactored PromptHistoryViewer

v.3.4.18 (Apr 20, 2025)

  • Changed edit icon for clarity

  • Added argument support to FunctionManager

  • Added save button to CodeGen

  • Fixed multi-image rendering in IconGen

  • Improved ModelSelector by hiding providers with no models

v.3.4.17 (Apr 20, 2025)

  • Fixed index out of range issue in VoiceSelectorGUI

v.3.4.16 (Apr 18, 2025)

  • Added FunctionManager

  • [Demo] Realtime API + Function integration

v.3.4.15 (Apr 18, 2025)

  • Fixed model categorizing regex

  • Added cancel request feature to GENTask

  • Implemented IconGenWindow

  • Cleaned up codegen files

  • Added output path tracking for GeneratedImage and GeneratedAudio

  • Fixed realtime API sample voice paths

  • Fixed Addressables amsdef

v.3.4.14 (Apr 18, 2025)

  • Removed implicit operator from EPrefs to prevent Mono JIT crash

v.3.4.13 (Apr 17, 2025)

  • Added missing constraints to:

    • Samples.Editor

    • MaterialDesign

    • MaterialDesign.Editor

  • Moved Setup folder outside of assembly

v.3.4.12 (Apr 17, 2025)

  • Fixed ElevenLabs voice library errors

v.3.4.08 (Apr 17, 2025)

  • Refactored EditorChatToolbar as separate component

  • Implemented export chat as .txt

  • Displayed chat list names as date instead of thread ID

  • Added more options to GENText and ElevenLabs GENTasks

  • Added Texture generator

  • Added default model settings to each API

  • Added fallback model support using default models

  • Renamed ImageModel → DallEModel, TTSModel → OpenAITTS

  • Fixed issue fetching Google models

  • Improved model family categorization

v.3.4.07 (Apr 16, 2025)

  • Fixed TTS file path issues

  • Fixed build errors related to StartOnInit

  • Fixed Unity hang on startup

  • Fixed model update triggering every time

  • Added ElevenLabs subscription feature

  • Added VoiceChanger and AudioIsolation to GENTask

  • Added more options to GENText

  • Improved EditorChat GUI and stream handling

  • Added drag & drop support in EditorChat

  • Added AI-powered script menu options:

    • "Improve Readability"

    • "Refactor"

v.3.4.16 (Apr 18, 2025)

  • Added model categorizing regex

  • Added cancel request feature to GENTask

  • New feature: IconGenWindow

  • Added output path tracking to GeneratedImage and GeneratedAudio

  • Fixed Addressables amsdef

v.3.4.2 (Apr 13, 2025)

  • Broken GUIDs fixed

v.3.2.0 (Apr 10, 2025)

  • Compile errors resolved

  • Refactored tons of codes

v.3.1.4 (Apr 7, 2025)

  • Build errors resolved

v.3.1.2 (Mar 31, 2025)

  • Added Google image models (Gemini, Imagen) to Editor Tool

  • Added new Google model families

  • Refactored the project

v.3.0.0 (Mar 17, 2025)

  • Added automatic model updates on Editor startup

  • Added automatic model enum code generation

PreviousUpdate LogsNextAI Dev Kit v2

Last updated