MiMo-V2.6-Flash

Name: MiMo-V2.6-Flash
Author: Xiaomi

Multimodal

Xiaomi

MiMo-V2.6-Flash is Xiaomi's MIT-licensed omnimodal sparse Mixture-of-Experts model with 309B total parameters and 15B active parameters. It supports text, image, video, and audio understanding with a 1,048,576-token context window and text output. MiMo-V2.6 scales reinforcement learning across coding, general agents, visual tasks, and cybersecurity. Xiaomi's launch article is dated September 22, 2026; released weights are named MiMo-V2.6-Flash-RL. Benchmark results are Xiaomi-reported release evaluations, distinct from intermediate live-training checkpoints.

Key Specifications

Parameters

309.0B

Context

1.0M

Release Date

September 22, 2026

Average Score

55.6%

API Documentation Research Paper Model Weights Results Blog

Timeline

Key dates in the model's history

Announcement / Last Update

September 22, 2026

Today

October 5, 2026

Technical Specifications

Parameters

309.0B

Training Tokens

Knowledge Cutoff

Family

Capabilities

MultimodalZeroEval

Pricing & Availability

Input (per 1M tokens)

$0.14

Output (per 1M tokens)

$0.28

Max Input Tokens

1.0M

Max Output Tokens

Supported Features

Function CallingStructured OutputCode ExecutionWeb SearchBatch InferenceFine-tuning

Benchmark Results

Model performance metrics across various tests and benchmarks

Other Tests

Specialized benchmarks

Agents' Last Exam

Agents' Last Exam; MiMo-V2.6-Flash. Published score divided by 100. • Self-reported

27.6%

AutomationBench v1.0.6

Automation Bench v1.0.6; MiMo-V2.6-Flash. Published score divided by 100. • Self-reported

52.3%

CyberGym

CyberGym; MiMo-V2.6-Flash. Published score divided by 100. • Self-reported

95.1%

DeepSWE 1.1

DeepSWE v1.1; MiMo-V2.6-Flash. Published score divided by 100. • Self-reported

67.9%

ExploitBench

ExploitBench; MiMo-V2.6-Flash. Published score divided by 100. • Self-reported

25.3%

ExploitGym

ExploitGym; MiMo-V2.6-Flash. Published score divided by 100. • Self-reported

6.0%

JobBench

JobBench; MiMo-V2.6-Flash. Published score divided by 100. • Self-reported

61.2%

MiMo Coding Bench

MiMo Code Bench; MiMo-V2.6-Flash. Published score divided by 100. • Self-reported

61.2%

MiMo Cyber Bench

MiMo Cyber Bench; MiMo-V2.6-Flash. Published score divided by 100. • Self-reported

77.2%

MiMo Visual Coding

MiMo Visual Coding; MiMo-V2.6-Flash. Published score divided by 100. • Self-reported

71.5%

OSWorld-Verified

OSWorld-Verified; MiMo-V2.6-Flash. Published score divided by 100. • Self-reported

80.8%

Program Bench

ProgramBench; MiMo-V2.6-Flash. Published score divided by 100. • Self-reported

26.0%

SEC-bench Pro

SEC Bench Pro; MiMo-V2.6-Flash. Published score divided by 100. • Self-reported

47.5%

Terminal-Bench 2.1

Terminal Bench 2.1; MiMo-V2.6-Flash. Published score divided by 100. • Self-reported

87.6%

Terminal-Bench 4.0

Terminal Bench 4.0; MiMo-V2.6-Flash. Published score divided by 100. • Self-reported

28.8%

Toolathlon-Verified

Toolathlon-verified; MiMo-V2.6-Flash. Published score divided by 100. • Self-reported

73.6%

License & Metadata

License

mit

Announcement Date

September 22, 2026

Last Updated

September 22, 2026

Compare MiMo-V2.6-Flash

All comparisons

vs MiMo-V2.6-Pro vs IBM Granite 4.2 3B vs LFM2.5-2.6B vs Claude Opus 5 vs Grok 4.7 vs IBM Granite 4.2 8B vs GPT-6 Sol vs Laguna S 2.1

Similar Models

All Models

MiMo-V2.6-Pro

Xiaomi

MM1.0T

Released:Sep 2026

Price:$0.43/1M tokens

MiMo-V2.5

Xiaomi

MM310.8B

Released:Apr 2026

Price:$0.17/1M tokens

MiMo-V2-Omni

Xiaomi

Released:Mar 2026

Price:$0.40/1M tokens

MiMo-V2-Flash

Xiaomi

309.0B

Best score:0.8 (GPQA)

Released:Dec 2025

MiMo-V2.5-Pro

Xiaomi

1.0T

Best score:1.0 (ARC)

Released:Apr 2026

Price:$0.43/1M tokens

MiniMax M3

MiniMax

MM428.0B

Released:Jun 2026

Price:$0.30/1M tokens

MAI-Code-1.1-Flash

Microsoft

MM138.0B

Released:Aug 2026

Price:$0.20/1M tokens

Mistral Medium 3.5

Mistral AI

MM128.0B

Released:Apr 2026

Price:$1.50/1M tokens

Recommendations are based on similarity of characteristics: developer organization, multimodality, parameter size, and benchmark performance. Choose a model to compare or go to the full catalog to browse all available AI models.