Xiaomi
Released September 21, 2026Cutoff December 2024

MiMo-V2.6-Pro

MiMo-V2.6-Pro is Xiaomi's trillion-parameter omni-modal reasoning model, trained by scaling RL on verifiable tasks, with a 1M-token context window.

Inputs
Text
Image
Audio
Video
Outputs
Text

Model Overview

Capabilities, design details, and architectural traits

MiMo-V2.6-Pro - Xiaomi's trillion-parameter omni-modal flagship

MiMo-V2.6-Pro is Xiaomi's flagship reasoning model. It is natively omni-modal, jointly understanding images, video, audio, and text, and it is fully open-sourced. It is built for complex projects, long-horizon tasks, high-stakes work, cybersecurity, and research.

TraitDetail
CreatorXiaomi
IdentityTrillion-parameter, omni-modal flagship reasoning model, released open-source
Training approachRSI path: scaling RL compute on verifiable, complex tasks so the model expands its capability frontier through exploration and feedback
RL run30 RL steps over roughly 750k trajectories in under six days, with the production run streamed live as it happened
Context window1M tokens input, 128K tokens max output
CapabilitiesDeep thinking, tool call, streaming, web search, structured output, context caching
Collaboration styleTuned for multimodal, multi-agent, and multi-harness collaboration; breaks tasks into sub-tasks and coordinates multiple agents
Target workflowsCoding, office work, design, research, content creation, cybersecurity, and computer operation

Built in public through RL

What separates MiMo-V2.6-Pro from most flagship releases is how it was trained. Xiaomi describes its approach as the RSI path, scaling reinforcement learning compute on verifiable, complex tasks rather than relying on standard post-training alone. The full production run was streamed live, and Xiaomi reports that RL proved sample-efficient and generalized beyond the training distribution.

From coding to co-scientist

Xiaomi highlights workflows where the model acts as a coordinator rather than a single responder. In its 3D open-world game demo, it decomposes a prompt into sub-tasks and directs multiple agents to build scenes, interaction logic, and visual checks. In materials research, it works as a co-scientist, reviewing literature, forming hypotheses, and simulating binding strength for candidate MOF materials.

Benchmark Performance

Independent evaluations · Artificial Analysis

46.3%
Intelligence

Accuracy & Capability Details

Humanity's Last Exam49.4%
SciCode - Scientific Coding60.9%
Long Context Reasoning86.3%

Compare Models Side-by-Side

Evaluate specifications, pricing, and independent benchmark indices

Model Details
General Info
ProviderXiaomiAnthropicAnthropic
Release DateSeptember 21, 2026September 22, 2026September 1, 2026
Knowledge CutoffDec 2024--
Context & Limits
Context Window1M1M1M
Pricing (per 1M tokens)
Input Pricing
$0.43
Best Input Pricing
$4$10
Output Pricing
$0.87
Best Output Pricing
$20$50
Modalities
Inputs
textimageaudiovideo
textimagefile
textimagefile
Outputs
text
text
text
Benchmarks (0-100)
Intelligence Index46.3
57.6
Best Intelligence Index
53.4
Coding Index--81.6
Agentic Index--57.9
MiMo-V2.6-Pro
Claude Opus 5.5
Claude Fable 5.1

Humanity's Last Exam

Extremely difficult logical reasoning and knowledge.

49%
MiMo-V2.6-Pro
Humanity's Last Exam
Score: 49%
MiMo-V2.6-Pro
61%
Claude Opus 5.5
Humanity's Last Exam
Score: 61%
Claude Opus 5.5
59%
Claude Fable 5.1
Humanity's Last Exam
Score: 59%
Claude Fable 5.1

Long Context Reasoning

Logical reasoning over long context windows.

86%
MiMo-V2.6-Pro
Long Context Reasoning
Score: 86%
MiMo-V2.6-Pro
85%
Claude Opus 5.5
Long Context Reasoning
Score: 85%
Claude Opus 5.5
85%
Claude Fable 5.1
Long Context Reasoning
Score: 85%
Claude Fable 5.1

SciCode Benchmark

Scientific coding and mathematical modeling.

61%
MiMo-V2.6-Pro
SciCode Benchmark
Score: 61%
MiMo-V2.6-Pro
67%
Claude Opus 5.5
SciCode Benchmark
Score: 67%
Claude Opus 5.5
63%
Claude Fable 5.1
SciCode Benchmark
Score: 63%
Claude Fable 5.1

Independent evaluation data provided by Artificial Analysis. To view the latest benchmarks and full details, visit their official site.