Skip to main content
In AI Providers, you can view all the models and storages available to your project organized by their type, and choose default providers or models for each type. To open AI Providers, in Settings, select AI Providers.
Available configurations depend on the selected project. To view or manage project-specific configurations, select the project in the project switcher.
AI Providers Your rights to view and edit AI Providers depend on your role and permissions. Default selections apply at the project level.

​
Available AI Providers

ELITEA supports these AI provider types:

LLMs

Manage language model configurations and choose default, high-tier, and low-tier models.

Embedding Models

Manage embedding models used for indexing, semantic search, and retrieval.

Vector Storage

Manage vector storage configurations used to store and retrieve embeddings.

Image Generation

Manage image generation models and choose the default model for image creation.

Speech Recognition (ASR)

Manage speech recognition models used for server-side voice input.

Text to Speech (TTS)

Manage text-to-speech models used for spoken playback of AI responses.

AI Credentials

Manage credentials used by AI providers and models.
For each AI provider configuration, you can see:
  • Provider icon
  • Configuration name
  • Status that shows whether the configuration is local or shared
  • Default or tier badges, when applicable
If you have permission to update configurations, you can open a configuration card and edit its details. AI Provider configuration cards

​
Default Models

Default models control which configurations ELITEA selects automatically for the current project.
  • Default model is used for most activities.
  • High-tier model is used for more complex work.
  • Low-tier model is used for routine tasks.
  • Default embedding model is used for embedding operations.
  • Default vector storage is used for vector database operations.
  • Default image generation model is used for image generation tasks.
  • Default ASR model is used for server-side voice input.
  • Default TTS model is used for spoken playback.
Changes to default selections apply at the project level.

​
LLMs

LLMs displays configured Large Language Model providers and their models. LLMs
Each LLM configuration card displays:
  • Provider icon
  • Model name
  • Status that shows whether the configuration is local or shared
  • Tier badges for High-tier, Low-tier, or Default, when applicable
  • Status that shows whether the configuration can be edited
Default Model Settings To set the default, high-tier, or low-tier model, select the necessary model from the corresponding dropdown list: Set default LLM Edit LLMs (Authorized Only) Users with configuration update permissions can open an LLM configuration card and edit its details. To edit an LLM model:
  1. Ensure that its credentials exist.
  2. Select the necessary LLM configuration card.
  3. Change the required fields:
  4. Save the configuration.
The model name must exactly match the provider’s identifier. Refer to your provider’s documentation:
  • OpenAI: Model names
  • Anthropic: Claude models
  • Azure OpenAI: Deployment name from your Azure portal
  • Vertex AI: Model names from Google Cloud console

​
Embedding Models

Embedding models convert text into numerical vector representations (embeddings) that capture semantic meaning. These vectors enable similarity comparisons, semantic search, and retrieval-augmented generation (RAG) workflows. Embedding models are essential for features like document indexing, knowledge base search, and context retrieval. Common examples include OpenAI, Azure OpenAI, HuggingFace, and Vertex AI embedding models. Embedding Models
Each embedding model configuration card displays:
  • Provider icon
  • Model name
  • Status that shows whether the configuration is local or shared
  • Status that shows whether the configuration can be edited
Edit Embedding Models (Authorized Only) Users with configuration update permissions can open an embedding model configuration card and edit its details. To edit an embedding model:
  1. Ensure that its credentials exist.
  2. Select the necessary embedding model configuration card.
  3. Change the required fields:
  4. Save the configuration.
After you save the model, ELITEA makes it available among other embedding models.

​
Vector Storages

Vector storage systems (vector databases) store and manage embedding vectors generated by embedding models. These specialized databases enable efficient similarity searches across large collections of embeddings, supporting features like semantic search, document retrieval, and knowledge base queries. Vector databases optimize storage and retrieval of high-dimensional vector data. Default vector storage specifies where embeddings are stored for search or retrieval. Choose it based on persistence and scale requirements. Vector Storages
Each vector storage configuration card displays:
  • Provider icon
  • Storage name
  • Icon that shows whether the configuration is local or shared
Edit Vector Storages To edit a vector storage:
  1. Select the necessary vector storage configuration card.
  2. Choose PGVector from the available configuration types
  3. Change the required fields:
  4. Save the configuration.
After you save the model, ELITEA makes it available among other vector storages.
For PGVector configurations, ensure the PostgreSQL database has the pgvector extension installed:

​
Image Generation

Image generation models create images from text descriptions (text-to-image) or modify existing images based on prompts. These AI models enable creative workflows, visual content creation, and automated image production based on natural language instructions. Default image generation model specifies the default model for image generation tasks. Image Generation
Each image generation model configuration card displays:
  • Provider icon
  • Model name
  • Status that shows whether the configuration is local or shared
  • Status that shows whether the configuration can be edited
Edit Image Generation Models (Authorized Only) Users with configuration update permissions can open an image generation model configuration card and edit its details. To edit an image generation model:
  1. Ensure that its credentials exist.
  2. Select the necessary image generation model configuration card.
  3. Change the required fields:
  4. Save the configuration.
After you save the model, ELITEA makes it available among other image generation models.

​
Speech Recognition

Speech Recognition (ASR) models convert spoken audio into text. A configured ASR model powers the server-side Voice Input feature in Chat, Agents, and Pipelines — providing real-time transcription as users speak. When no ASR model is configured, Voice Input falls back to the browser’s built-in Web Speech API. If neither is available, the microphone icon is hidden. Default ASR model specifies the default model used for server-side voice transcription across Chat, Agents, and Pipelines. If no ASR model is configured, ELITEA falls back to the browser Web Speech API. If neither option is available, ELITEA hides the microphone control. Speech Recognition ASR Modes ELITEA determines the ASR mode from the model name. The model name determines the transcription mode. Use names such as whisper-1 or gpt-4o-transcribe for batch transcription.
Each speech recognition model configuration card displays:
  • Provider icon
  • Model name
  • Status that shows whether the configuration is local or shared
  • Status that shows whether the configuration can be edited
Edit Speech Recognition Models (Authorized Only) Users with configuration update permissions can open a speech recognition model configuration card and edit its details. To edit a ASR model:
  1. Ensure that its credentials exist.
  2. Select the necessary speech recognition model configuration card.
  3. Change the required fields:
  4. Save the configuration.
After you save the model, ELITEA makes it available among other speech recognition models.

​
Text to Speech Models

Text to Speech (TTS) models synthesize spoken audio from text. A configured TTS model powers the server-side Text-to-Speech and Speaking Mode features in Chat, Agents, and Pipelines — reading AI responses aloud using Web Audio API playback. When no TTS model is configured, playback falls back to the browser’s built-in SpeechSynthesis API. Default TTS model specifies the default model used for reading AI responses aloud across Chat, Agents, and Pipelines. If no TTS model is configured, ELITEA falls back to the browser SpeechSynthesis API. Text to Speech
Each text-to-speech model configuration card displays:
  • Provider icon
  • Model name
  • Status that shows whether the configuration is local or shared
  • Status that shows whether the configuration can be edited
To edit a TTS model:
  1. Ensure that its credentials exist.
  2. Select the necessary text-to-speech model configuration card.
  3. Change the required fields:
  4. Save the configuration.
After you save the model, ELITEA makes it available among other text-to-speech models.
Server-side TTS delivers audio as PCM-16-LE chunks via Socket.IO, played back through the Web Audio API at 24 kHz. Text is highlighted in sync with playback. Pause and resume are fully supported for both server-side and browser-fallback TTS.

​
AI Credentials

The AI Credentials section only appears when credentials are configured in your project. If no credentials exist, this section will be hidden.
AI Credentials are authentication configurations that store API keys, tokens, endpoints, and connection strings for accessing AI service providers. These credentials serve as the foundation for other configurations, enabling secure communication with external AI platforms without exposing sensitive authentication details in individual model configurations. One credential configuration can support multiple models. AI Credentials
Each AI credential configuration card displays:
  • Provider icon
  • Credential name
  • Status that shows whether the configuration is local or shared
To add new AI credentials:
  1. Create a new AI credential configuration.
  2. Choose the provider you want to configure.
  3. Enter the provider-specific authentication details (see below).
  4. Save the configuration.

​
Available AI Service Providers

Provider-Specific Fields: LLM Models Use an LLM Model configuration to register a model that references one of the AI credentials above. The exact Name value must match the provider identifier or Azure deployment name used by the selected credential.
AI Dial and Reasoning ModelsWhen you create an OpenAI or Anthropic LLM model and use AI Dial as the selected AI Credentials provider, do not enable Supports Reasoning.OpenAI and Anthropic model configurations that use AI Dial credentials must have Supports Reasoning turned off.AI Dial and Reasoning Models
For OpenAI: For Azure OpenAI: For Vertex AI: For Amazon Bedrock: For AI Dial: For Ollama:

​
Troubleshooting

If you cannot find or use the expected configuration, check the current project, your permissions, and the default configuration for the relevant section.
If a section or configuration does not appear:
  1. Check that you selected the correct project in the project switcher.
  2. Refresh Settings > AI Providers.
  3. Confirm that the project contains configurations for that section.
Available configurations are project-specific. If the selected project does not contain configurations for a section, ELITEA does not show configuration cards for that section.
If you can open AI Providers but cannot edit a configuration:
  1. Open the configuration card.
  2. Check whether ELITEA shows a disabled edit state.
  3. Ask a project administrator to grant configuration update permissions if needed.
ELITEA allows editing only when your role includes configuration update permissions.
If voice input does not use a server-side speech recognition model:
  1. In Settings > AI Providers, check the Speech Recognition (ASR) section.
  2. Confirm that the project has at least one ASR configuration.
  3. Confirm that a default ASR model is selected.
  4. Check your browser connection.
If no ASR model is configured, ELITEA falls back to the browser Web Speech API. If neither the server model nor the browser API is available, ELITEA hides the microphone control.
If spoken playback does not use a server-side TTS model:
  1. In Settings > AI Providers, check the Text to Speech (TTS) section.
  2. Confirm that the project has at least one TTS configuration.
  3. Confirm that a default TTS model is selected.
If no TTS model is configured, ELITEA falls back to the browser SpeechSynthesis API.
If an ASR model uses the wrong transcription mode, check the model name.ELITEA determines the ASR mode from the model name:
  • Names that include whisper or transcribe use batch transcription.
  • Names that do not include whisper or transcribe use the realtime streaming path.
If a provider configuration does not work after you save it:
  1. Open the configuration and review the provider settings.
  2. Confirm that the selected credentials are correct for that provider.
  3. Save the configuration again after you correct the values.
ELITEA supports connection validation for configuration types that expose connection checks.
If an OpenAI or Anthropic LLM model fails connection validation when it uses AI Dial credentials:
  1. Open the LLM model configuration.
  2. Check the AI Credentials field and confirm that it uses AI Dial.
  3. Clear the Supports Reasoning option.
  4. Save the configuration again.
  5. Run the connection check again.
When AI Dial is used as the credential provider for OpenAI or Anthropic LLM models, connection validation fails if Supports Reasoning is enabled. Keep Supports Reasoning disabled for these configurations.