The v3.0.4 default,
ibm/granite-20b-code-instruct, has been withdrawn from multitenant watsonx.ai. Set id to a model available in your deployment. See IBM model lifecycle updates.Parameters
| Parameter | Type | Default | Description |
|---|---|---|---|
id | str | "ibm/granite-20b-code-instruct" | The id of the IBM WatsonX model to use |
name | str | "WatsonX" | The name of the model |
provider | str | "IBM" | The provider of the model |
frequency_penalty | Optional[float] | None | Penalizes new tokens based on their frequency in the text so far |
presence_penalty | Optional[float] | None | Penalizes new tokens based on whether they appear in the text so far |
max_tokens | Optional[int] | None | Maximum number of tokens to generate |
temperature | Optional[float] | None | Controls randomness in the model’s output |
top_p | Optional[float] | None | Controls diversity via nucleus sampling |
logprobs | Optional[int] | None | Return log probabilities of the output tokens |
top_logprobs | Optional[int] | None | Number of most likely tokens to return log probabilities for |
request_params | Optional[Dict[str, Any]] | None | Additional parameters to include in the request |
api_key | Optional[str] | None | The API key for IBM WatsonX (defaults to IBM_WATSONX_API_KEY env var) |
project_id | Optional[str] | None | The project ID for IBM WatsonX (defaults to IBM_WATSONX_PROJECT_ID env var) |
url | Optional[str] | "https://eu-de.ml.cloud.ibm.com" | The WatsonX endpoint URL (IBM_WATSONX_URL env var takes precedence) |
verify | bool | True | Verify TLS certificates when connecting |
client_params | Optional[Dict[str, Any]] | None | Declared client parameters. Agno v3.0.4 does not pass this field to ModelInference; provide model_client for custom client configuration |
model_client | Optional[ModelInference] | None | Pre-configured WatsonX ModelInference client |
model_type | ModelType | ModelType.MODEL | Functional role of this model (e.g. MODEL, OUTPUT_MODEL, PARSER_MODEL). Set by the agent |
supports_native_structured_outputs | bool | False | Whether the model supports structured outputs natively |
supports_json_schema_outputs | bool | False | Whether the model uses JSON schema for structured outputs |
system_prompt | Optional[str] | None | System prompt from the model added to the agent |
instructions | Optional[List[str]] | None | Instructions from the model added to the agent |
tool_message_role | str | "tool" | Role of tool messages |
assistant_message_role | str | "assistant" | Role of assistant messages |
cache_response | bool | False | Cache model responses to avoid redundant API calls during development |
cache_ttl | Optional[int] | None | Time-to-live for cached responses, in seconds |
cache_dir | Optional[str] | None | Directory for the response cache. Defaults to ~/.agno/cache/model_responses |
retries | int | 0 | Number of retries to attempt before raising a ModelProviderError |
delay_between_retries | int | 1 | Delay between retries, in seconds |
exponential_backoff | bool | False | If True, the delay between retries is doubled each time |
retry_with_guidance | bool | True | Retry a failed model invocation once with a guidance message appended |
retry_with_guidance_limit | int | 1 | Number of times to retry the model invocation with guidance |