None
None
Long Story Video Skill
Turn your idea into 10s–10min video—AI writes scripts, prompts, and generates footage automatically.
Ads Video Skill
Generate professional ads and sales videos—AI auto-generates scripts, prompts, and footage.
3D Science Explainer Video Skill
Convert scientific concepts into stunning 3D explain animations
Feedback
freeTrialImage.bannerPity
freeTrialImage.upgradeUnlock
- ✓freeTrialImage.benefitHd
- ✓freeTrialImage.benefitWatermark
- ✓freeTrialImage.benefitUnlimited
minimax m3.1 flash
MiniMax M3.1 Flash accelerates agent, coding and computer-use jobs, combining 200+ TPS output with roughly 70% lower token spend.
All Tools
Discover our comprehensive AI-powered animation toolkit
MiniMax H3
MiniMax H3 AI Video Generator
Seedance 2.5
The Future of AI Video Is Here.
Seedance 2.0
The Future of AI Video Is Here.
Veo3.1
Create Stunning Videos with Veo3.1
FLUX 3 Video Generator
Kling 3.0
Next-Gen AI Video Generator
Grok Video Generator
Create Videos from Text or Images with AI
MiniMax H3 video generator
MiniMax H3 AI Video Generator
MiniMax M3.1 Flash for Complex Agent and Code Work
Built from the anonymous Space Bunny system, MiniMax M3.1 Flash has been tested on enterprise automation, software engineering, computer operation and security workloads.
- Multi-Step Agent AutomationPosts top-tier AutomationBench-AA results and 76.2 on BankerToolBench, completing SaaS workflows across sales, marketing, operations, service, finance and HR.
- Code Comprehension and RepairScores 73.7% on DeepSWE v1.1, handling code reading, feature development, bug fixes and test-based validation over long task chains.
- Real-World Computer Operation and SecurityRecords 61.7 on OSWorld V2 and 29.6 on Agents' Last Exam for live-environment control, plus 82% on CyberGym for finding and verifying vulnerabilities.
Three Steps to Start with MiniMax M3.1 Flash
Move from initial setup to production-ready agent workloads in three practical steps.
Capabilities Built into MiniMax M3.1 Flash
Quicker generation, predictable low token pricing, and proven performance in automation, coding, computer use and security engineering.
Sparser Serving for Higher Speed
Deeper sparsity in the serving stack pushes everyday output past 200 TPS, with P50 staying above 170 TPS, making online inference nearly three times faster.
Simple, Low Token Rates
$0.15 per 1M input tokens, $0.60 per 1M output tokens and $0.015 per 1M cached input reduce overall agentic spend by roughly 70% versus the previous model.
SaaS Workflow Execution at Scale
A leading AutomationBench-AA position and 76.2 on BankerToolBench show multi-step planning, financial compliance and linked tool calls inside live SaaS environments.
Software Development on DeepSWE
A 73.7% score on DeepSWE v1.1 puts it near leading models on hard software tasks that need independent code reading, feature building, defect repair and test checks.
Computer Use in Real Environments
Scores of 61.7 on OSWorld V2 and 29.6 on Agents' Last Exam, first among Chinese models on both, demonstrate tool use and task delivery in live environments.
Security Engineering Assistance
An 82% CyberGym score, supported by internal expert review, spans code auditing, vulnerability patching, penetration testing, malware analysis and reverse engineering at low cost.
MiniMax M3.1 Flash Questions Answered
Find answers on MiniMax M3.1 Flash pricing, output speed, benchmark results and access options.
What is MiniMax M3.1 Flash?
It is MiniMax’s newer Flash model, developed from the anonymous Space Bunny system that handled more than 200 trillion tokens, with advances in agent work, coding, computer use and cybersecurity.
What does it cost per million tokens?
Pricing is flat: $0.15 per 1M input tokens, $0.60 per 1M output tokens and $0.015 per 1M cached input, making overall agentic use roughly 70% less expensive than the prior model.
Why is inference nearly three times faster?
Greater sparsity in the serving stack lifts routine output to 200+ TPS while keeping median (P50) speed above 170 TPS.
Which benchmarks support these results?
It posts a top-tier AutomationBench-AA position, 76.2 on BankerToolBench, 73.7% on DeepSWE v1.1, 61.7 on OSWorld V2, 29.6 on Agents' Last Exam and 82% on CyberGym.
How can I access the model?
Use the MiniMax API at platform.minimax.io, an M Plan subscription or MiniMax Code, with the desktop build available from agent.minimax.io/download; weights are planned later.
How does Flash compare with the Pro model?
Flash continues MiniMax’s work on balancing efficiency and capability, while Pro, focused on extending intelligence limits, is scheduled for a later release.
Put MiniMax M3.1 Flash to Work
Pick the MiniMax API, an M Plan subscription or MiniMax Code to put demanding agent, code, computer-use and security jobs into production, with roughly 70% lower spend and nearly 3x faster inference.