MaaS_Mimo_V2.6

MaaS_Mimo_V2.6_pro_20260922

Model Overview

item	value
Model ID (model)	mimo-v2.6-pro
Positioning	All-modality flagship, designed for professional, long-haul and complex workflows
context window	1M tokens
Maximum Output	128K tokens
Input Modality	text, images, videos, and audio
output modality	Text
Default Thinking Mode	Enable (for Chat / Messages, thinking. typedefaults to enabled; for Responses, reasoning. effortdefaults to enabled)
competence	Full-modal understanding, deep thinking, tool calling, streaming output, structured output, context caching

Request Protocol

https

Header

Parameter Name	Type	Required	Description
Content-Type	string	Yes	application/json
Authorization	string	Yes	Bearer ${your_AK}

OpenAI Chat Completions

Request URL

POST https://genaiapi.cloudsway.net/v1/ai/{endpointPath}/chat/completions

Request Body Parameters

Parameter Name	Secondary Parameters	Tertiary Parameters	Level-4 Parameters	Type	Required	Description
model	-	-	-	string	Yes	Fixed as MaaS_Mimo_V2.6_pro_20260922 (original model from the vendor: mimo-v2.6-pro)
messages	-	-	-	array[json]	Yes	Conversation message list. The source document lists five types of message objects: Developer, System, User, Assistant and Tool
	role	-	-	string	Yes	Role. Optional values for Developer messages: developer; the FAQ specifies that the OpenAI protocol includes developer / system / user / assistant
	content	-	-	string or array	Yes	Message content. A string indicates plain text; an array represents multimodal content parts
		type	-	string	No	content part type. Text: text; Image: image_url; Audio: input_audio; Video: video_url
		text	-	string	No	Text content
		image_url	url	string	No	Public URL of the image, or data:{MIME_TYPE};base64,$BASE64_IMAGE. Supported formats: JPEG / PNG / GIF / WebP / BMP; both the URL and Base64 data must not exceed 50 MB
		input_audio	data	string	No	Public URL of the audio, or data:{MIME_TYPE};base64,$BASE64_AUDIO. Supported formats: MP3 / WAV / FLAC / M4A / OGG; URL ≤ 100 MB, Base64 ≤ 50 MB
		video_url	url	string	No	Public URL of the video, or data:{MIME_TYPE};base64,$BASE64_VIDEO. Supported formats: MP4 / MOV / AVI / WMV; URL ≤ 300 MB, Base64 ≤ 50 MB
		fps	-	number	No	Video frame extraction frame rate, which controls the timing precision. Defaults to 2, with a range of[0.1,10]
		media_resolution	-	string	No	Video single frame resolution level. Default default;maxenhances detail recognition
	name	-	-	string	No	Participant optional name, used to distinguish identical roles
	reasoning_content	-	-	string	No	Assistant's historical thinking content. During multi-round tool calls in thinking mode, it is recommended to retain all history reasoning_content
frequency_penalty	-	-	-	number | null	No	Penalize new tokens based on their occurrence frequency in the existing text. Defaults to 0, with a valid range of[-2.0,2.0]
max_completion_tokens	-	-	-	integer | null	No	Maximum visible output + inference tokens. This model defaults to 131072, with a valid range of[1,131072]
presence_penalty	-	-	-	number | null	No	Depending on whether a penalty for new tokens has already been applied. Defaults to 0, with a valid range of[-2.0,2.0]
response_format	-	-	-	object	No	Specify the output format
	type	-	-	string	Yes	textis the default text;json_objectenables JSON mode. JSON mode only guarantees syntactic validity, and the field structure is constrained by the prompt.
stop	-	-	-	string | array | null	No	Up to 4 stop sequences; the returned text will not contain these sequences. Defaults to null
stream	-	-	-	boolean | null	No	true, results will be returned via SSE streaming. Defaults to false
thinking	-	-	-	object	No	Controls whether to enable chain-of-thought. Custom settings are not supported in thinking mode for temperature/ top_p, and any values passed will be forcibly set to 1.0/ 0.95
	type	-	-	string	Yes	This model is by default enabled. Accepted values: enabled, disabled
temperature	-	-	-	number	No	Sampling temperature, range[0,1.5]. The default value for this model is 1.0. It is recommended to modify only one of top_pand this parameter. This value is forced to 1.0 in thinking mode.
tool_choice	-	-	-	string	No	Controls how tools are selected. Accepted values: auto. If a value other than autois passed, the backend will remove this field, and the behavior will remain equivalent to auto
tools	-	-	-	array	No	List of callable tools. Supports function tools and web search tools
	type	-	-	string	Yes	The function tool is function; the web search tool is web_search
	function	name	-	string	Yes	Function name. a-zA-Z0-9_-, length 1-64
	function	description	-	string	No	Function function description
	function	parameters	-	object	No	JSON Schema. Omission indicates an empty parameter list
	function	strict	-	boolean	No	Whether to strictly follow parameters. Defaults to false
	max_keyword	-	-	integer	No	Web search: the maximum number of keywords per search round, which is used to control the number of calls and costs. The example value in the source document is 3
	force_search	-	-	boolean	No	Web Search: trueForce web search. If not forced, the model will determine whether to perform a search
	limit	-	-	integer	No	Web search: Source document example appears, meaning not specified separately
	user_location	type	-	string	No	Perform a web search for the geographic location. An example from the source document is approximate
	user_location	country	-	string	No	Country, source document example China
	user_location	region	-	string	No	Region, Source Document Example Hubei
	user_location	city	-	string	No	City, Source Document Example Wuhan
top_p	-	-	-	number	No	Nuclear sampling, range[0.01,1.0]. Default value 0.95. Forced to 0.95 in thinking mode.

Request Example

curl -X POST 'https://genaiapi.cloudsway.net/v1/ai/{endpointPath}/chat/completions' \
  -H 'Content-Type: application/json' \
  -H 'Authorization: Bearer ${your_AK}' \
  -d '{
    "model": "MaaS_Mimo_V2.6_pro_20260922",
    "messages": [
      {
        "role": "system",
        "content": "You are MiMo, an AI assistant developed by Xiaomi."
      },
      {
        "role": "user",
        "content": "please introduce yourself"
      }
    ],
    "max_completion_tokens": 1024,
    "temperature": 1.0,
    "top_p": 0.95,
    "stream": false,
    "thinking": {
      "type": "disabled"
    }
  }'

Response Example

{
  "id": "6272d55f-a84e-4664-9db1-8462899db7dc_abdfaf015d1a4623b2a4f9b99179c2f9",
  "choices": [
    {
      "finish_reason": "stop",
      "index": 0,
      "message": {
        "content": "Hi there! I'm MiMo, a friendly AI assistant created by the Xiaomi LLM Core Team. Nice to meet you! How can I help you today?",
        "role": "assistant",
        "tool_calls": null
      }
    }
  ],
  "created": 1790006669,
  "model": "MaaS_Mimo_V2.6_pro_20260922",
  "object": "chat.completion",
  "usage": {
    "completion_tokens": 34,
    "prompt_tokens": 57,
    "total_tokens": 91,
    "completion_tokens_details": {
      "reasoning_tokens": 0
    },
    "prompt_tokens_details": {
      "cached_tokens": 0
    }
  }
}

Example of Streaming Request

curl -X POST 'https://genaiapi.cloudsway.net/v1/ai/{endpointPath}/chat/completions' \
  -H 'Content-Type: application/json' \
  -H 'Authorization: Bearer ${your_AK}' \
  -d '{
    "model": "MaaS_Mimo_V2.6_pro_20260922",
    "messages": [
      {"role": "user", "content": "please introduce yourself"}
    ],
    "max_completion_tokens": 1024,
    "stream": true,
    "thinking": {"type": "disabled"}
  }'

SDK Calling Method

import os
from openai import OpenAI

client = OpenAI(
    ${your_AK}=os.environ.get("your_AK"),
    base_url="https://genaiapi.cloudsway.net/v1/ai/{endpointPath}"
)

completion = client.chat.completions.create(
    model="MaaS_Mimo_V2.6_pro_20260922",
    messages=[
        {"role": "system", "content": "You are MiMo, an AI assistant developed by Xiaomi."},
        {"role": "user", "content": "please introduce yourself"}
    ],
    max_completion_tokens=1024,
    stream=False,
    extra_body={"thinking": {"type": "disabled"}}
)
print(completion.model_dump_json())

OpenAI Responses

Request URL

POST https://genaiapi.cloudsway.net/v1/ai/{endpoint}/responses

Request Body Parameters

Parameter Name	Secondary Parameters	Tertiary Parameters	Type	Required	Description
model	-	-	string	Yes	Fixed as MaaS_Mimo_V2.6_pro_20260922 (original model from the vendor: mimo-v2.6-pro)
input	-	-	string or array	Yes	Text / image / audio / video input. A string is equivalent to user text; an array is of type InputItemList
	type	-	string	No	The text input type expanded from the source document is input_text. Additionally, ResponseInputImage/ ResponseInputAudioContent/ ResponseInputVideoContentare listed, with their subfields not expanded.
	text	-	string	No	Text input content
instructions	-	-	string	No	System (or developer) instructions injected into the model context
max_output_tokens	-	-	integer	No	Maximum visible output + inference tokens. This model defaults to 131072, with a valid range of[1,131072]
stream	-	-	boolean	No	true, results will be returned via SSE streaming. Defaults to false
reasoning	-	-	object	No	Inference Configuration. In the thinking mode, multi-round tool calls shall retain historical thinking content in subsequent inputfields
	effort	-	string	Yes	Control thinking investment. nonedisables thinking; all other valid values enable thinking with consistent behavior, with no intensity differentiation for now. minimalmaps to low;xhigh/ max/ ultramap to high. This model defaults to enabled. Available values: none, minimal, low, medium, high, xhigh, max, ultra
temperature	-	-	number	No	Sampling temperature, range [0, 1.5]. This model defaults to 1.0. Thinking mode under forced 1.0
text	-	-	object	No	Text response configuration, supporting plain text or structured JSON
	format	type	string	Yes	; additionally, text; additionally, ResponseFormatJSONObjectis listed separately. The default value is{"type":"text"}
tool_choice	-	-	string	No	Optional values: auto. Any value other than autowill be removed by the backend, and the behavior will remain the same as auto
tools	-	-	array	No	Callable tools. The source document lists Function / Custom / Namespace, and only expands the Function section.
	type	-	string	Yes	The Function tool is function
	name	-	string	Yes	Function name. a-zA-Z0-9_-, length 1-64
	parameters	-	object	Yes	Function Parameter JSON Schema
	strict	-	boolean	Yes	Strictly follow the parameter mode. Default value: false
	description	-	string | null	No	function description
top_p	-	-	number	No	Nuclear sampling, range[0.01,1.0]. Default value 0.95. Forcibly set to 0.95 in thinking mode

Request Example

curl -X POST 'https://genaiapi.cloudsway.net/v1/ai/{endpoint}/responses' \
  -H 'Content-Type: application/json' \
  -H 'Authorization: Bearer ${your_AK}' \
  -d '{
    "model": "MaaS_Mimo_V2.6_pro_20260922",
    "instructions": "You are MiMo, an AI assistant developed by Xiaomi.",
    "input": "please introduce yourself",
    "max_output_tokens": 1024,
    "stream": false,
    "reasoning": {
      "effort": "none"
    }
  }'

Response Example

{
  "id": "resp_bcdb1b61-d49e-48e5-8289-384ad1e65f2f_9aa9e9dfd5b84cb99b088fe4e53b65ec",
  "object": "response",
  "created_at": 1790007468,
  "status": "completed",
  "error": null,
  "incomplete_details": null,
  "model": "MaaS_Mimo_V2.6_pro_20260922",
  "metadata": null,
  "output": [
    {
      "id": "msg_3d0b3a6faf634d36a8a5bdb8f2100fcd",
      "type": "message",
      "status": "completed",
      "role": "assistant",
      "content": [
        {
          "type": "output_text",
          "text": "Hey there! I'm MiMo, Xiaomi's AI assistant.",
          "annotations": []
        }
      ]
    }
  ],
  "output_text": "Hey there! I'm MiMo, Xiaomi's AI assistant.",
  "usage": {
    "input_tokens": 57,
    "input_tokens_details": {
      "cached_tokens": 0
    },
    "output_tokens": 130,
    "output_tokens_details": {
      "reasoning_tokens": 0
    },
    "total_tokens": 187
  }
}

Example of streaming request

curl -X POST 'https://genaiapi.cloudsway.net/v1/ai/{endpoint}/responses' \
  -H 'Content-Type: application/json' \
  -H 'Authorization: Bearer ${your_AK}' \
  -d '{
    "model": "MaaS_Mimo_V2.6_pro_20260922",
    "input": "please introduce yourself",
    "stream": true,
    "reasoning": {"effort": "none"}
  }'

SDK Calling Method

import os
from openai import OpenAI

client = OpenAI(
    ${your_AK}=os.environ.get("${your_AK}"),
    base_url="https://genaiapi.cloudsway.net/v1/ai/{endpoint}"
)

response = client.responses.create(
    model="MaaS_Mimo_V2.6_pro_20260922",
    instructions="You are MiMo, an AI assistant developed by Xiaomi.",
    input="please introduce yourself",
    max_output_tokens=1024,
    stream=False,
    extra_body={"reasoning": {"effort": "none"}}
)
print(response.model_dump_json())

Anthropic Messages

Request URL

POST https://genaiapi.cloudsway.net/{endpoint}/v1/messages

Request Body Parameters

Parameter Name	Secondary Parameters	Tertiary Parameters	Level-4 Parameters	Type	Required	Description
model	-	-	-	string	Yes	Fixed as MaaS_Mimo_V2.6_pro_20260922 (original model from the vendor: mimo-v2.6-pro)
messages	-	-	-	array[json]	Yes	Enter messages. Each must contain roleand content. If the last message is from the assistant, the response will continue from this content.
	role	-	-	string	Yes	Valid values: user, assistant, system
	content	-	-	string or array	Yes	A string is an abbreviation for a single text; an array refers to content blocks
		type	-	string	No	Text is text; image is image. There are separate audio/video example entries on the protocol page, and the field details are not expanded.
		text	-	string	No	Text content
		source	type	string	No	Image source. urlor base64
		source	url	string	No	Public URL of the image
		source	media_type	string	No	Base64 image MIME, e. g. image/png
		source	data	string	No	Pure Base64 image data (excluding the data URL prefix)
max_tokens	-	-	-	integer	No	Maximum tokens to generate before stopping. This model defaults to 131072, with a range of[1,131072]. The model may stop early.
stop_sequences	-	-	-	array	No	Custom stop sequence. Upon natural termination, stop_reasonis end_turn
stream	-	-	-	boolean	No	Whether to enable streaming output. Defaults to false
system	-	-	-	string or array	No	System prompt, which can be text or content parts
temperature	-	-	-	number	No	Sampling temperature, range[0,1.5]. The default value for this model is 1.0. It is forced to be 1.0 in thinking mode.
thinking	-	-	-	object	No	Expand thinking configuration: multi-round tool calls in thinking mode should retain the history thinkingcontent blocks
	type	-	-	string	Yes	This model is by default enabled. Accepted values: enabled, disabled
tool_choice	-	-	-	object	No	Control how to use the tools
	type	-	-	string	Yes	Optional values: auto. Non- autovalues will be removed by the backend
	disable_parallel_tool_use	-	-	boolean	No	trueand type= auto, at most one tool call will be output. Defaults to false
tools	-	-	-	array	No	Tool definition. The model may return tool_use, and the caller performs postback of tool_result
	name	-	-	string	Yes	Tool name, corresponding to the name in tool_use
	description	-	-	string	No	Tool description, please provide as much detail as possible
	type	-	-	string	No	Optional values: custom
	input_schema	type	-	string	Yes	Only for object
	input_schema	properties	-	object | null	No	Tool Input Attributes
	input_schema	required	-	array | null	No	List of Required Attributes
top_p	-	-	-	number	No	Nucleus sampling, range[0.01,1.0]. Default value 0.95. Forced to 0.95in thinking mode. It is recommended to adjust only either temperatureor top_p

The source document does not require additional anthropic-versionheaders.

Request Example

curl -X POST 'https://genaiapi.cloudsway.net/{endpoint}/v1/messages' \
  -H 'Content-Type: application/json' \
  -H 'Authorization: Bearer ${your_AK}' \
  -d '{
    "model": "MaaS_Mimo_V2.6_pro_20260922",
    "max_tokens": 1024,
    "system": "You are MiMo, an AI assistant developed by Xiaomi.",
    "messages": [
      {
        "role": "user",
        "content": [
          {"type": "text", "text": "please introduce yourself"}
        ]
      }
    ],
    "top_p": 0.95,
    "stream": false,
    "temperature": 1.0,
    "thinking": {"type": "disabled"}
  }'

Response Example

{
  "id": "23894237-e793-4156-965e-56cea4295290_a43d65ad5fb048e7ba57e27e689b2bf8",
  "type": "message",
  "role": "assistant",
  "model": "MaaS_Mimo_V2.6_pro_20260922",
  "stop_reason": "end_turn",
  "content": [
    {
      "type": "text",
      "text": "Hey there! I'm MiMo, Xiaomi's AI assistant, here to help you with whatever you need."
    }
  ],
  "usage": {
    "input_tokens": 57,
    "output_tokens": 41
  }
}

Example of streaming request

curl -X POST 'https://genaiapi.cloudsway.net/{endpoint}/v1/messages' \
  -H 'Content-Type: application/json' \
  -H 'Authorization: Bearer ${your_AK}' \
  -d '{
    "model": "MaaS_Mimo_V2.6_pro_20260922",
    "max_tokens": 1024,
    "messages": [
      {"role": "user", "content": "please introduce yourself"}
    ],
    "stream": true,
    "thinking": {"type": "disabled"}
  }'

SDK Calling Method

import os
from anthropic import Anthropic

client = Anthropic(
    ${your_AK}=os.environ.get("${your_AK}"),
    base_url="https://genaiapi.cloudsway.net/{endpoint}/v1"
)

message = client.messages.create(
    model="MaaS_Mimo_V2.6_pro_20260922",
    max_tokens=1024,
    system="You are MiMo, an AI assistant developed by Xiaomi.",
    messages=[
        {
            "role": "user",
            "content": [{"type": "text", "text": "please introduce yourself"}]
        }
    ],
    top_p=0.95,
    stream=False,
    temperature=1.0
)
print(message.content)

MaaS_Mimo_V2.6_flash_20260922

Model Overview

item	value
Model ID (model)	mimo-v2.6-flash
Positioning	All-modal efficient inference for professional office high-frequency calls and large-scale tasks
context window	1M tokens
Maximum Output	128K tokens
Input Modality	text, images, videos, and audio
output modality	Text
Default Thinking Mode	Enable (for Chat / Messages, thinking. typedefaults to enabled; for Responses, reasoning. effortdefaults to enabled)
competence	Full-modal understanding, deep reasoning, tool calling, streaming output, internet search (Chat Completions only), structured output, context caching
Context Caching	Support; response usagereturns cached_tokens/ cache_read_input_tokens. No separate request parameters are provided in the source document.

Request Protocol

https

Header

Parameter Name	Type	Required	Description
Content-Type	string	Yes	application/json
Authorization	string	Yes	Bearer ${your_AK}

OpenAI Chat Completions

Request URL

POST https://genaiapi.cloudsway.net/v1/ai/{endpointPath}/chat/completions

Request Body Parameters

Parameter Name	Secondary Parameters	Tertiary Parameters	Level-4 Parameters	Type	Required	Description
model	-	-	-	string	Yes	Fixed as MaaS_Mimo_V2.6_flash_20260922 (original model from vendor: mimo-v2.6-flash)
messages	-	-	-	array[json]	Yes	Conversation message list. The source document lists five types of message objects: Developer, System, User, Assistant and Tool
	role	-	-	string	Yes	Role. Optional values for Developer messages: developer; the FAQ specifies that the OpenAI protocol includes developer / system / user / assistant
	content	-	-	string or array	Yes	Message content. A string indicates plain text; an array represents multimodal content parts
		type	-	string	No	content part type. Text: text; Image: image_url; Audio: input_audio; Video: video_url
		text	-	string	No	Text content
		image_url	url	string	No	Public URL of the image, or data:{MIME_TYPE};base64,$BASE64_IMAGE. Supported formats: JPEG / PNG / GIF / WebP / BMP; both the URL and Base64 data must not exceed 50 MB
		input_audio	data	string	No	Public URL of the audio, or data:{MIME_TYPE};base64,$BASE64_AUDIO. Supported formats: MP3 / WAV / FLAC / M4A / OGG; URL ≤ 100 MB, Base64 ≤ 50 MB
		video_url	url	string	No	Public URL of the video, or data:{MIME_TYPE};base64,$BASE64_VIDEO. Supported formats: MP4 / MOV / AVI / WMV; URL ≤ 300 MB, Base64 ≤ 50 MB
		fps	-	number	No	Video frame extraction frame rate, which controls the timing precision. Defaults to 2, with a range of[0.1,10]
		media_resolution	-	string	No	Video single frame resolution level. Default default;maxenhances detail recognition
	name	-	-	string	No	Participant optional name, used to distinguish identical roles
	reasoning_content	-	-	string	No	Assistant's historical thinking content. During multi-round tool calls in thinking mode, it is recommended to retain all history reasoning_content
frequency_penalty	-	-	-	number | null	No	Penalize new tokens based on their occurrence frequency in the existing text. Defaults to 0, with a valid range of[-2.0,2.0]
max_completion_tokens	-	-	-	integer | null	No	Maximum visible output + inference tokens. This model defaults to 131072, with a valid range of[1,131072]
presence_penalty	-	-	-	number | null	No	Depending on whether a penalty for new tokens has already been applied. Defaults to 0, with a valid range of[-2.0,2.0]
response_format	-	-	-	object	No	Specify the output format
	type	-	-	string	Yes	textis the default text;json_objectenables JSON mode. JSON mode only guarantees syntactic validity, and the field structure is constrained by the prompt.
stop	-	-	-	string | array | null	No	Up to 4 stop sequences; the returned text will not contain these sequences. Defaults to null
stream	-	-	-	boolean | null	No	true, results will be returned via SSE streaming. Defaults to false
thinking	-	-	-	object	No	Controls whether to enable chain-of-thought. Custom settings are not supported in thinking mode for temperature/ top_p, and any values passed will be forcibly set to 1.0/ 0.95
	type	-	-	string	Yes	This model is by default enabled. Accepted values: enabled, disabled
temperature	-	-	-	number	No	Sampling temperature, range[0,1.5]. The default value for this model is 1.0. It is recommended to modify only one of top_pand this parameter. This value is forced to 1.0 in thinking mode.
tool_choice	-	-	-	string	No	Controls how tools are selected. Accepted values: auto. If a value other than autois passed, the backend will remove this field, and the behavior will remain equivalent to auto
tools	-	-	-	array	No	List of callable tools. Supports function tools and web search tools
	type	-	-	string	Yes	The function tool is function; the web search tool is web_search
	function	name	-	string	Yes	Function name. a-zA-Z0-9_-, length 1-64
	function	description	-	string	No	Function function description
	function	parameters	-	object	No	JSON Schema. Omission indicates an empty parameter list
	function	strict	-	boolean	No	Whether to strictly follow parameters. Defaults to false
	max_keyword	-	-	integer	No	Web search: the maximum number of keywords per search round, which is used to control the number of calls and costs. The example value in the source document is 3
	force_search	-	-	boolean	No	Web Search: trueForce web search. If not forced, the model will determine whether to perform a search
	limit	-	-	integer	No	Web search: Source document example appears, meaning not specified separately
	user_location	type	-	string	No	Perform a web search for the geographic location. An example from the source document is approximate
	user_location	country	-	string	No	Country, source document example China
	user_location	region	-	string	No	Region, Source Document Example Hubei
	user_location	city	-	string	No	City, Source Document Example Wuhan
top_p	-	-	-	number	No	Nuclear sampling, range[0.01,1.0]. Default value 0.95. Forcibly set to 0.95 in thinking mode

Request Example

curl -X POST 'https://genaiapi.cloudsway.net/v1/ai/{endpointPath}/chat/completions' \
  -H 'Content-Type: application/json' \
  -H 'Authorization: Bearer ${your_AK}' \
  -d '{
    "model": "MaaS_Mimo_V2.6_flash_20260922",
    "messages": [
      {
        "role": "system",
        "content": "You are MiMo, an AI assistant developed by Xiaomi."
      },
      {
        "role": "user",
        "content": "please introduce yourself"
      }
    ],
    "max_completion_tokens": 1024,
    "temperature": 1.0,
    "top_p": 0.95,
    "stream": false,
    "thinking": {
      "type": "disabled"
    }
  }'

Response Example

{
  "id": "6272d55f-a84e-4664-9db1-8462899db7dc_abdfaf015d1a4623b2a4f9b99179c2f9",
  "choices": [
    {
      "finish_reason": "stop",
      "index": 0,
      "message": {
        "content": "Hi there! I'm MiMo, a friendly AI assistant created by the Xiaomi LLM Core Team. Nice to meet you! How can I help you today?",
        "role": "assistant",
        "tool_calls": null
      }
    }
  ],
  "created": 1790006669,
  "model": "MaaS_Mimo_V2.6_flash_20260922",
  "object": "chat.completion",
  "usage": {
    "completion_tokens": 34,
    "prompt_tokens": 57,
    "total_tokens": 91,
    "completion_tokens_details": {
      "reasoning_tokens": 0
    },
    "prompt_tokens_details": {
      "cached_tokens": 0
    }
  }
}

Example of streaming request

curl -X POST 'https://genaiapi.cloudsway.net/v1/ai/{endpointPath}/chat/completions' \
  -H 'Content-Type: application/json' \
  -H 'Authorization: Bearer ${your_AK}' \
  -d '{
    "model": "MaaS_Mimo_V2.6_flash_20260922",
    "messages": [
      {"role": "user", "content": "please introduce yourself"}
    ],
    "max_completion_tokens": 1024,
    "stream": true,
    "thinking": {"type": "disabled"}
  }'

SDK Calling Method

import os
from openai import OpenAI

client = OpenAI(
    ${your_AK}=os.environ.get("${your_AK}"),
    base_url="https://genaiapi.cloudsway.net/v1/ai/{endpointPath}"
)

completion = client.chat.completions.create(
    model="MaaS_Mimo_V2.6_flash_20260922",
    messages=[
        {"role": "system", "content": "You are MiMo, an AI assistant developed by Xiaomi."},
        {"role": "user", "content": "please introduce yourself"}
    ],
    max_completion_tokens=1024,
    stream=False,
    extra_body={"thinking": {"type": "disabled"}}
)
print(completion.model_dump_json())

OpenAI Responses

Request URL

POST https://genaiapi.cloudsway.net/v1/ai/{endpoint}/responses

Request Body Parameters

Parameter Name	Secondary Parameters	Tertiary Parameters	Type	Required	Description
model	-	-	string	Yes	Fixed as MaaS_Mimo_V2.6_flash_20260922 (original model from the manufacturer: mimo-v2.6-flash)
input	-	-	string or array	Yes	Text / image / audio / video input. A string is equivalent to user text; an array is of type InputItemList
	type	-	string	No	The text input type expanded from the source document is input_text. Additionally, ResponseInputImage/ ResponseInputAudioContent/ ResponseInputVideoContentare listed, with their subfields not expanded.
	text	-	string	No	Text input content
instructions	-	-	string	No	System (or developer) instructions injected into the model context
max_output_tokens	-	-	integer	No	Maximum visible output + inference tokens. This model defaults to 131072, with a valid range of[1,131072]
stream	-	-	boolean	No	true, results will be returned via SSE streaming. Defaults to false
reasoning	-	-	object	No	Inference Configuration. In the thinking mode, multi-round tool calls shall retain historical thinking content in subsequent inputfields
	effort	-	string	Yes	Control thinking investment. nonedisables thinking; all other valid values enable thinking with consistent behavior, with no intensity differentiation for now. minimalmaps to low;xhigh/ max/ ultramap to high. This model defaults to enabled. Available values: none, minimal, low, medium, high, xhigh, max, ultra
temperature	-	-	number	No	Sampling temperature, range[0,1.5]. The default value for this model is 1.0. It is forced to be 1.0 in thinking mode.
text	-	-	object	No	Text response configuration, supporting plain text or structured JSON
	format	type	string	Yes	; additionally, text; additionally, ResponseFormatJSONObjectis listed separately. The default value is{"type":"text"}
tool_choice	-	-	string	No	Optional values: auto. Any value other than autowill be removed by the backend, and the behavior will remain the same as auto
tools	-	-	array	No	Callable tools. The source document lists Function / Custom / Namespace, and only expands the Function section.
	type	-	string	Yes	The Function tool is function
	name	-	string	Yes	Function name. a-zA-Z0-9_-, length 1-64
	parameters	-	object	Yes	Function Parameter JSON Schema
	strict	-	boolean	Yes	Strictly follow the parameter mode. Default value: false
	description	-	string | null	No	function description
top_p	-	-	number	No	Nuclear sampling, range[0.01,1.0]. Default value 0.95. Forcibly set to 0.95 in thinking mode

Request Example

curl -X POST 'https://genaiapi.cloudsway.net/v1/ai/{endpoint}/responses' \
  -H 'Content-Type: application/json' \
  -H 'Authorization: Bearer ${your_AK}' \
  -d '{
    "model": "MaaS_Mimo_V2.6_flash_20260922",
    "instructions": "You are MiMo, an AI assistant developed by Xiaomi.",
    "input": "please introduce yourself",
    "max_output_tokens": 1024,
    "stream": false,
    "reasoning": {
      "effort": "none"
    }
  }'

Response Example

{
  "id": "resp_bcdb1b61-d49e-48e5-8289-384ad1e65f2f_9aa9e9dfd5b84cb99b088fe4e53b65ec",
  "object": "response",
  "created_at": 1790007468,
  "status": "completed",
  "error": null,
  "incomplete_details": null,
  "model": "MaaS_Mimo_V2.6_flash_20260922",
  "metadata": null,
  "output": [
    {
      "id": "msg_3d0b3a6faf634d36a8a5bdb8f2100fcd",
      "type": "message",
      "status": "completed",
      "role": "assistant",
      "content": [
        {
          "type": "output_text",
          "text": "Hey there! I'm MiMo, Xiaomi's AI assistant.",
          "annotations": []
        }
      ]
    }
  ],
  "output_text": "Hey there! I'm MiMo, Xiaomi's AI assistant.",
  "usage": {
    "input_tokens": 57,
    "input_tokens_details": {
      "cached_tokens": 0
    },
    "output_tokens": 130,
    "output_tokens_details": {
      "reasoning_tokens": 0
    },
    "total_tokens": 187
  }
}

Example of streaming request

curl -X POST 'https://genaiapi.cloudsway.net/v1/ai/{endpoint}/responses' \
  -H 'Content-Type: application/json' \
  -H 'Authorization: Bearer ${your_AK}' \
  -d '{
    "model": "MaaS_Mimo_V2.6_flash_20260922",
    "input": "please introduce yourself",
    "stream": true,
    "reasoning": {"effort": "none"}
  }'

SDK Calling Method

import os
from openai import OpenAI

client = OpenAI(
    ${your_AK}=os.environ.get("${your_AK}"),
    base_url="https://genaiapi.cloudsway.net/v1/ai/{endpoint}"
)

response = client.responses.create(
    model="MaaS_Mimo_V2.6_flash_20260922",
    instructions="You are MiMo, an AI assistant developed by Xiaomi.",
    input="please introduce yourself",
    max_output_tokens=1024,
    stream=False,
    extra_body={"reasoning": {"effort": "none"}}
)
print(response.model_dump_json())

Anthropic Messages

Request URL

POST https://genaiapi.cloudsway.net/{endpoint}/v1/messages

Request Body Parameters

Parameter Name	Secondary Parameters	Tertiary Parameters	Level-4 Parameters	Type	Required	Description
model	-	-	-	string	Yes	Fixed as MaaS_Mimo_V2.6_flash_20260922 (original manufacturer model: mimo-v2.6-flash)
messages	-	-	-	array[json]	Yes	Enter messages. Each must contain roleand content. If the last message is from the assistant, the response will continue from this content.
	role	-	-	string	Yes	Valid values: user, assistant, system
	content	-	-	string or array	Yes	A string is an abbreviation for a single text; an array refers to content blocks
		type	-	string	No	Text is text; image is image. There are separate audio/video example entries on the protocol page, and the field details are not expanded.
		text	-	string	No	Text content
		source	type	string	No	Image source. urlor base64
		source	url	string	No	Public URL of the image
		source	media_type	string	No	Base64 image MIME, e. g. image/png
		source	data	string	No	Pure Base64 image data (excluding the data URL prefix)
max_tokens	-	-	-	integer	No	Maximum tokens to generate before stopping. This model defaults to 131072, with a range of[1,131072]. The model may stop early.
stop_sequences	-	-	-	array	No	Custom stop sequence. Upon natural termination, stop_reasonis end_turn
stream	-	-	-	boolean	No	Whether to enable streaming output. Defaults to false
system	-	-	-	string or array	No	System prompt, which can be text or content parts
temperature	-	-	-	number	No	Sampling temperature, range[0,1.5]. The default value for this model is 1.0. It is forced to be 1.0 in thinking mode.
thinking	-	-	-	object	No	Expand thinking configuration: multi-round tool calls in thinking mode should retain the history thinkingcontent blocks
	type	-	-	string	Yes	This model is by default enabled. Accepted values: enabled, disabled
tool_choice	-	-	-	object	No	Control how to use the tools
	type	-	-	string	Yes	Optional values: auto. Non- autovalues will be removed by the backend
	disable_parallel_tool_use	-	-	boolean	No	trueand type= auto, at most one tool call will be output. Defaults to false
tools	-	-	-	array	No	Tool definition. The model may return tool_use, and the caller performs postback of tool_result
	name	-	-	string	Yes	Tool name, corresponding to the name in tool_use
	description	-	-	string	No	Tool description, please provide as much detail as possible
	type	-	-	string	No	Optional values: custom
	input_schema	type	-	string	Yes	Only for object
	input_schema	properties	-	object | null	No	Tool Input Attributes
	input_schema	required	-	array | null	No	List of Required Attributes
top_p	-	-	-	number	No	Nucleus sampling, range[0.01,1.0]. Default value 0.95. Forced to 0.95in thinking mode. It is recommended to adjust only either temperatureor top_p

Request Example

curl -X POST 'https://genaiapi.cloudsway.net/{endpoint}/v1/messages' \
  -H 'Content-Type: application/json' \
  -H 'Authorization: Bearer ${your_AK}' \
  -d '{
    "model": "MaaS_Mimo_V2.6_flash_20260922",
    "max_tokens": 1024,
    "system": "You are MiMo, an AI assistant developed by Xiaomi.",
    "messages": [
      {
        "role": "user",
        "content": [
          {"type": "text", "text": "please introduce yourself"}
        ]
      }
    ],
    "top_p": 0.95,
    "stream": false,
    "temperature": 1.0,
    "thinking": {"type": "disabled"}
  }'

Response Example

{
  "id": "23894237-e793-4156-965e-56cea4295290_a43d65ad5fb048e7ba57e27e689b2bf8",
  "type": "message",
  "role": "assistant",
  "model": "MaaS_Mimo_V2.6_flash_20260922",
  "stop_reason": "end_turn",
  "content": [
    {
      "type": "text",
      "text": "Hey there! I'm MiMo, Xiaomi's AI assistant, here to help you with whatever you need."
    }
  ],
  "usage": {
    "input_tokens": 57,
    "output_tokens": 41
  }
}

Example of streaming request

curl -X POST 'https://genaiapi.cloudsway.net/{endpoint}/v1/messages' \
  -H 'Content-Type: application/json' \
  -H 'Authorization: Bearer ${your_AK}' \
  -d '{
    "model": "MaaS_Mimo_V2.6_flash_20260922",
    "max_tokens": 1024,
    "messages": [
      {"role": "user", "content": "please introduce yourself"}
    ],
    "stream": true,
    "thinking": {"type": "disabled"}
  }'

SDK Calling Method

import os
from anthropic import Anthropic

client = Anthropic(
    ${your_AK}=os.environ.get("${your_AK}"),
    base_url="https://genaiapi.cloudsway.net/{endpoint}/v1"
)

message = client.messages.create(
    model="MaaS_Mimo_V2.6_flash_20260922",
    max_tokens=1024,
    system="You are MiMo, an AI assistant developed by Xiaomi.",
    messages=[
        {
            "role": "user",
            "content": [{"type": "text", "text": "please introduce yourself"}]
        }
    ],
    top_p=0.95,
    stream=False,
    temperature=1.0
)
print(message.content)