Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.agno.com/llms.txt

Use this file to discover all available pages before exploring further.

Xiaomi MiMo serves its models through an OpenAI-compatible API, so you drive them through Agno like any other OpenAI-compatible provider. The MiMo class defaults to mimo-v2.5-pro and points at https://api.xiaomimimo.com/v1.

​
Authentication

Sign in with a Xiaomi account (register at id.mi.com if you don’t have one), then create a key in the console under API Keys. Set your MIMO_API_KEY environment variable. 
export MIMO_API_KEY=***

​
Example

Use MiMo with your Agent: 
from agno.agent import Agent
from agno.models.xiaomi import MiMo

agent = Agent(model=MiMo(id="mimo-v2.5-pro"), markdown=True)

# Print the response in the terminal
agent.print_response("Share a 2 sentence horror story.")
View more examples here.

​
Parameters

ParameterTypeDefaultDescription
idstr"mimo-v2.5-pro"The id of the MiMo model to use
namestr"MiMo"The name of the model
providerstr"Xiaomi MiMo"The provider of the model
api_keyOptional[str]NoneThe API key for MiMo (defaults to MIMO_API_KEY env var)
base_urlstr"https://api.xiaomimimo.com/v1"The base URL for the MiMo API
use_thinkingOptional[bool]NoneControl thinking mode. See Thinking mode.
MiMo extends the OpenAI-compatible interface and supports most parameters from the OpenAI model. Note: MiMo supports JSON mode but not native json_schema structured outputs, so supports_native_structured_outputs is set to False. Use use_json_mode=True for structured output.

​
Thinking mode

Control thinking mode with the use_thinking flag:
ValueBehavior
None (default)The flag is not sent; the API uses the model default.
TrueForce thinking on. The model returns reasoning_content.
FalseForce thinking off for a faster, cheaper response.
from agno.agent import Agent
from agno.models.xiaomi import MiMo

agent = Agent(model=MiMo(id="mimo-v2.5-pro", use_thinking=True), markdown=True)