Get Capabilities

GET

/api/v1/chat/capabilities

Discover chat system capabilities.

Returns available LLM models, embedding models with index status, supported context modes, retrieval strategies, languages, and defaults.

Use this endpoint to populate UI dropdowns and validate parameters before making chat requests.

Args: config: Application configuration (injected) index_service: Index management service (injected)

Returns: CapabilitiesResponse with all available options and defaults

Responses

200

Successful Response

CapabilitiesResponse

Response model for chat capabilities discovery.

object

llm_models

Llm Models

Available LLM models

Array<object>

LLMModelInfo

Information about an available LLM model.

object

name

required

Name

Model identifier (profile_name or model)

string

display_name

required

Display Name

Human-readable name

string

provider

required

Provider

Provider type (openai, azure, anthropic, etc.)

string

model

Any of:

string

string

null

null

embedding_models

Embedding Models

Available embedding models

Array<object>

EmbeddingModelInfo

Information about an available embedding model.

object

name

required

Name

Model identifier

string

display_name

required

Display Name

Human-readable name

string

dimension

Any of:

integer

integer

null

null

vector_count

Vector Count

Number of vectors indexed

integer

0

has_index

Has Index

Whether index exists

boolean

index_type

Any of:

string

string

null

null

context_modes

Context Modes

Available context modes

Array<string>

default: rag,attachment,hybrid,document,none

retrieval_strategies

Retrieval Strategies

Available retrieval strategies

Array<string>

default: merge,sequential,ensemble

search_languages

Search Languages

Available FTS languages

Array<string>

default: norwegian,english,simple

defaults

ChatDefaults

Default parameter values

object

llm_model

Llm Model

Default LLM model

string

default: gpt-4o-mini

embedding_model

Embedding Model

Default embedding model

string

default: openai-small

context_mode

Context Mode

Default context mode

string

default: rag

retrieval_strategy

Retrieval Strategy

Default retrieval strategy

string

default: merge

temperature

Temperature

Default temperature

number

default: 0.7 <= 2

max_tokens

Max Tokens

Default max tokens

integer

default: 1024 >= 1 <= 8192

similarity_threshold

Similarity Threshold

Default similarity

number

default: 0.7 <= 1

max_results

Max Results

Default max results

integer

default: 10 >= 1 <= 50

search_language

Search Language

Default FTS language

string

default: norwegian