Speed
Output tokens / second
MiMo-V2.5 overview
MiMo-V2.5 by Xiaomi
MiMo-V2.5 is Xiaomi's native omnimodal AI, trained from scratch to understand text, images, audio, and video together.
MiMo-V2.5 specs
- Model ID
xiaomi/mimo-v2.5- Developer
- Xiaomi
- Context window
- 1,000,000 tokens
- Max output
- 128,000 tokens
- Input
- Text, Image
- Output
- Text
- API
- OpenAI-compatible, base URL
https://api.aimlapi.com/v1
What is MiMo-V2.5 API
Released on April 22, 2026 by Xiaomi's AI team, MiMo-V2.5 picks up where MiMo-V2-Omni left off, with substantially better agentic performance, sharper visual reasoning, and native support for a 1 million token context window. It's designed for production, not benchmarks alone.
The model sits at the base of the V2.5 family. For teams that need the absolute ceiling on agentic tasks and software engineering, there's MiMo-V2.5-Pro. For general-purpose multimodal work at an honest price, MiMo-V2.5 is the practical starting point.
Native omnimodal architecture
Not a text model with adapters, all four modalities trained jointly from day one, so cross-modal reasoning is genuinely coherent.
Surpasses MiMo-V2-Pro on agentic tasks
The newer multimodal model beats the previous text-only flagship on general agentic benchmarks, at half the token cost.
1 million token context
No more chunking. Process entire research papers, long video transcripts, or sprawling codebases without losing coherence.
Agent-framework ready
Works natively with Claude Code, OpenCode, Kilo, and other popular scaffolds. Drop it in without rewiring your stack.
Stepped-up token efficiency
Strong multimodal results with significantly fewer tokens than closed-source competitors, which keeps production costs predictable.
Use cases
MiMo-V2.5 is documented as suited to visual reasoning, including precise perception and complex chart analysis. It is also described as suited to deep multimodal understanding across text, image, video, and audio.
It is documented for general agent scenarios, with native agent execution across browsing, comprehension, reasoning, and operation. Xiaomi also describes it as supporting long-range tasks such as long video tracking, lengthy document analysis, and extended temporal reasoning.
It is additionally presented as useful for everyday task performance in agent workflows. The model is also described as a native full-modal model for understanding and translating understanding into action.
Capabilities
- Function calling
- Streaming
- Structured outputs
- Reasoning
- Web search
MiMo-V2.5 benchmarks by 
Artificial Analysis
See how MiMo-V2.5 compares in generation speed and latency. Data retrieved Nov 3, 2026.
Benchmark results
Compare 11 models available on AI/ML API across 5 independent benchmarks.
Humanity's Last Exam
Expert-written questions across academic subjects.
GPQA Diamond
Graduate-level science questions written to resist web search.
Terminal-Bench 4.0
Agentic tasks completed in a command-line environment.
AA-LCR
Reasoning across long documents of about 100k tokens.
SciCode
Python coding tasks drawn from scientific research.
MiMo-V2.5 API pricing
| Usage | Price | Unit |
|---|---|---|
| Input | $0.1481 | 1M tokens |
| Output | $0.2962 | 1M tokens |
| Cached input | $0.003 | 1M tokens |
Prices from the AI/ML API price list, checked on Nov 4, 2026.
MiMo-V2.5 vs other models: token prices
AI/ML API input and output rates for 5 models, per million tokens.
| Model | Input | Output |
|---|---|---|
 MiMo-V2.5 | $0.1481 | $0.2962 |
| MiMo V2.6 Flash | $0.1481 | $0.2962 |
| MiMo V2.6 Pro | $0.4602 | $0.9205 |
| MiMo-V2.5 Pro | $0.4602 | $0.9205 |
| MiMo V2.6 Pro UltraSpeed | $4.6 | $9.2 |
How to use the MiMo-V2.5 API
View docs
Build with your agent
Give your agent the model details, code and documentation.
Paste it into your coding agent to build the integration.
1Get your API key
Create API keyexport AIMLAPI_KEY="8f2b4c7d9a1e6b3f…"2Make your first request
curl 'https://api.aimlapi.com/v1/chat/completions' \
-H "Authorization: Bearer $AIMLAPI_KEY" \
-H 'Content-Type: application/json' \
-d '{
"model": "xiaomi/mimo-v2.5",
"messages": [
{
"role": "user",
"content": "Write a haiku about the ocean."
}
]
}'import os
import requests
headers = {
"Authorization": f"Bearer {os.environ['AIMLAPI_KEY']}",
"Content-Type": "application/json",
}
payload = {
"model": "xiaomi/mimo-v2.5",
"messages": [
{
"role": "user",
"content": "Write a haiku about the ocean."
}
]
}
response = requests.post("https://api.aimlapi.com/v1/chat/completions", headers=headers, json=payload, timeout=120)
response.raise_for_status()
print(response.json())const response = await fetch("https://api.aimlapi.com/v1/chat/completions", {
method: "POST",
headers: {
Authorization: `Bearer ${process.env.AIMLAPI_KEY}`,
"Content-Type": "application/json",
},
body: JSON.stringify({
"model": "xiaomi/mimo-v2.5",
"messages": [
{
"role": "user",
"content": "Write a haiku about the ocean."
}
]
}),
});
if (!response.ok) throw new Error(`HTTP ${response.status}`);
console.log(await response.json());Models similar to MiMo-V2.5
Related articles
MiMo-V2.5 FAQ
What is the context window of MiMo-V2.5?
MiMo-V2.5 has a 1,000,000 tokens context window and can return up to 128,000 tokens.
What inputs and outputs does MiMo-V2.5 support?
MiMo-V2.5 takes image, text as input and returns text.
What is the model ID for MiMo-V2.5?
Use xiaomi/mimo-v2.5 as the model id. Requests go to https://api.aimlapi.com/v1/chat/completions.
When was MiMo-V2.5 released?
MiMo-V2.5 became available on April 23, 2026.
How much does MiMo-V2.5 cost?
MiMo-V2.5 costs $0.1481 input and $0.2962 output per 1M tokens on AI/ML API, billed as you go with no subscription.
Does MiMo-V2.5 support streaming?
Yes, MiMo-V2.5 can stream responses as they are generated.
Can MiMo-V2.5 read images?
Yes, MiMo-V2.5 accepts image input alongside text.
Who created MiMo-V2.5?
MiMo-V2.5 was built by Xiaomi.