OpenRouter 模型 ↗ · 2026-09-22
xiaomi/mimo-v2.6-pro
Model release OpenRouter 模型 Source language: Chinese
OpenRouter 于 2026-09-22 收录该模型(收录时间不等于模型正式发布日期);上下文:1,050,000 tokens;输入价:$0.4350/百万 tokens;输出价:$0.8700/百万 tokens;定位(来源描述):MiMo-V2.6-Pro is the flagship foundation model developed by Xiaomi.
Full text
Xiaomi: MiMo-V2.6-Pro
xiaomi/mimo-v2.6-pro
Model weights
Compare
MiMo-V2.6-Pro is the flagship foundation model developed by Xiaomi. Built at a scale of over 1T parameters, it is designed to push the ceiling of capability for the most demanding workloads. The model features a 1M-token context window and native multimodal capabilities. Optimized for agentic workflows, it delivers top-tier performance across coding, visual, general, and research scenarios, excelling at complex, long-horizon tasks with robust generalization across a diverse range of agent harnesses.
Modalities
In / out price
25% off
$0.3263 / $0.6525per 1M
Context
1.1M
Released
Sep 21, 2026
Compare
Providers
Different companies host the same model. OpenRouter routes your request to one of them based on the routing mode you pick: Balanced (price + speed), Nitro (fastest), or Floor (cheapest).
Latency / throughput
10% off
$0.43$0.3870$0.87$0.7830$0.0036$0.003240.99s34 tok/s
100.00%
$0.4350$0.8700$0.003603.72s31 tok/s
100.00%
25% off
$0.435$0.3263$0.87$0.6525$0.004$0.003006.62s24 tok/s
99.53%
$4.0000$8.0000$0.036000.56s140 tok/s
98.95%
$0.5220$1.0440$0.004321.86s31 tok/s
90.49%
ProviderInput / Output /M
Input $0.387 / Output $0.783
10% off
34 tok/s100.00% uptime
Input $0.435 / Output $0.87
31 tok/s100.00% uptime
Input $0.3263 / Output $0.6525
25% off
24 tok/s99.53% uptime
Input $4.00 / Output $8.00
140 tok/s98.95% uptime
Input $0.522 / Output $1.044
31 tok/s90.49% uptime
Pricing
The average price customers actually pay for this model, next to the prices providers post. Caching and discounts mean the price actually paid is often well below the listed one.
Weighted average
Input
$0.06208/M tokens
Output
$0.9817/M tokens
MiMo-V2.6-Pro — Price history
Price history
No effective price data for this selection
$0.0214$0.65294.3%15.7%
$0.0547$0.87088.2%67.7%
$0.1177$0.78270.2%15.5%
$0.3163$8.00092.9%1.1%
$0.3736$1.04028.6%0.0%
ProviderInput / Output /M
Token share
Effective in $0.0214 / Effective out $0.652
Token share
15.7%
Effective in $0.0547 / Effective out $0.870
Token share
67.7%
Effective in $0.1177 / Effective out $0.782
Token share
15.5%
Effective in $0.3163 / Effective out $8.000
Token share
1.1%
Effective in $0.3736 / Effective out $1.040
Token share
0.0%
Performance
Throughput is how fast the model writes (tokens per second — higher is better). Latency is total round-trip time (lower is better). TTFT is time-to-first-token — how long before you see anything appear (lower is better).
Throughput
140tok/s
P50, best across providers
Latency
0.56s
P50, best provider
Latency / throughput
AutoExacto benchmarks
GPQA DiamondTAU-BenchNovita91.1%78.8%Xiaomi90.5%78.0%auto-routing89.0%78.8%StreamLake89.8%77.2%DeepInfra88.9%77.8%
+2 more providers
Uptime
Uptime is the percentage of the past 3 days that at least one provider was responding to requests. Availability is the percentage of time that inference was successfully served. OpenRouter continuously monitors and uses the next-best provider when one returns an error.
Uptime (3d)The model was reachable. Request routed to a provider.
100.00%
Availability (3d)The model returned inference from any provider. Errors and empty responses count against it.
99.82%
Availability over the last 3 days
Last 72 hours
Availability 99.82%
3 Days Ago2 Days AgoYesterdayNow
Availability over the last 24 hours
OpenRouter Availability
99.84%
Without Routing
99.38%
When an error occurs in an upstream provider, we can recover by routing to another healthy provider, if your request filters allow it. You can access per-provider uptime data programmatically through the Endpoints API. Learn more about our load balancing and provider routing options.
Benchmarks
Scores on standardized evaluations and this model’s own outputs on shared benchmark prompts. Higher percentages are better — and each rank shows where this model lands among the models evaluated on that source.
Benchmark score summary for Xiaomi: MiMo-V2.6-Pro (Artificial Analysis, Design Arena and Vals AI)
SourceBenchmarkScore
Artificial AnalysisMiMo-V2.6-Pro Intelligence Index29.9
Artificial AnalysisMiMo-V2.6-Pro HLE49.4%
Artificial AnalysisMiMo-V2.6-Pro AA-LCR86.3%
Artificial AnalysisMiMo-V2.6-Pro GDPval-AA59.6%
Artificial AnalysisMiMo-V2.6-Pro CritPt26.6%
Artificial AnalysisMiMo-V2.6-Pro SciCode60.9%
Artificial AnalysisMiMo-V2.6-Pro Terminal-Bench 4.034.8%
Artificial AnalysisMiMo-V2.6-Pro AA-Omniscience Accuracy34.8%
Artificial AnalysisMiMo-V2.6-Pro AA-Omniscience Non-Hallucination Rate39.8%
Design ArenaMiMo-V2.6-Pro Agents Arena Webapps Elo1269
Design ArenaMiMo-V2.6-Pro Models Arena 3D Elo1321
Design ArenaMiMo-V2.6-Pro Models Arena Code Categories Elo1316
Design ArenaMiMo-V2.6-Pro Models Arena Data Visualization Elo1335
Design ArenaMiMo-V2.6-Pro Models Arena Game Development Elo1312
Design ArenaMiMo-V2.6-Pro Models Arena SVG Elo1290
Design ArenaMiMo-V2.6-Pro Models Arena UI Component Elo1332
Design ArenaMiMo-V2.6-Pro Models Arena Website Elo1311
Vals AIMiMo V2.6 Pro Access to Justice Benchmark70.0%
Vals AIMiMo V2.6 Pro Code Migration v1.148.8%
Vals AIMiMo V2.6 Pro CyberBench v1.172.9%
Vals AIMiMo V2.6 Pro EMB v1.167.3%
Vals AIMiMo V2.6 Pro Finance Agent v2.164.3%
Vals AIMiMo V2.6 Pro Vals Finance Index57.8%
Vals AIMiMo V2.6 Pro Financial Coaching Bench75.4%
Vals AIMiMo V2.6 Pro HLAB-Vals-1.113.8%
Vals AIMiMo V2.6 Pro Immigration67.5%
Vals AIMiMo V2.6 Pro IOI39.3%
Vals AIMiMo V2.6 Pro Vals Legal Index72.1%
Vals AIMiMo V2.6 Pro Legal Research Bench47.1%
Vals AIMiMo V2.6 Pro MedCode45.0%
Vals AIMiMo V2.6 Pro MedScribe88.3%
Vals AIMiMo V2.6 Pro MysteryMechanism15.3%
Vals AIMiMo V2.6 Pro ProgramBench0.5%
Vals AIMiMo V2.6 Pro ProofBench v1.170.0%
Vals AIMiMo V2.6 Pro Public Benefits Bench v1.168.9%
Vals AIMiMo V2.6 Pro SAGE84.4%
Vals AIMiMo V2.6 Pro SRE Bench3.1%
Vals AIMiMo V2.6 Pro SWE-bench Multimodal41.3%
Vals AIMiMo V2.6 Pro Tax Agent Bench32.4%
Vals AIMiMo V2.6 Pro Terminal-Bench 4.031.3%
Vals AIMiMo V2.6 Pro Terminal-Bench-Science2.9%
Vals AIMiMo V2.6 Pro Vals Index57.8%
Vals AIMiMo V2.6 Pro Vibe Code Bench 1-100 v1.126.4%
Vals AIMiMo V2.6 Pro Vibe Code Bench v1.185.2%
Apps
Public apps that send the most traffic to this model. Good signal for what real production workloads look like — and a hint at which use cases this model is best suited for.
1.
ISEKAI ZERO
AI adventures. Travel with your favorite characters
749Btokens
2.
Hermes Agent
Hermes Agent is an open-source, self-improving AI agent by Nous Research that runs persistently with memory across sessions, and builds reusable skills from experience. It comes with 40+ built-in tools, including web search, browser automation, and vision, plus scheduled automations and subagents.
523Btokens
3.
Claude Code
Claude Code is Anthropic's agentic coding tool that reads your entire codebase, plans and executes changes across files, runs tests, and iterates on failures, all from natural language prompts.
340Btokens
4.
Cline
Cline is an open-source AI coding agent that lives inside your IDE, autonomously exploring your codebase, editing files, running terminal commands, and using browser automation.
199Btokens
5.
omp
new
181Btokens
Top 5 Apps
5.96B
Rest of OpenRouter
5.04B
Activity
Token volume and request traffic to this model over time.
API
Drop-in code to call this model. OpenRouter's API is OpenAI-compatible — most SDKs work by just swapping the base URL. The only thing that changes between models is the model slug below.
Not currently available through EU in-region routing. In-region routing docs
Documentation
Guides, cookbooks, blog posts, and tools for building with Xiaomi: MiMo-V2.6-Pro on OpenRouter.
Open-Weight vs Closed-Weight Models: What's the Difference?Blog post
Frequently asked questions
MiMo-V2.6-Pro is the flagship foundation model developed by Xiaomi. Built at a scale of over 1T parameters, it is designed to push the ceiling of capability for the most demanding workloads. The model features a 1M-token context window and native multimodal capabilities.
The MiMo-V2.6-Pro documentation on OpenRouter includes the Open-Weight vs Closed-Weight Models: What's the Difference? blog post. The quickstart covers API keys and your first request, and the API reference lists every parameter.
MiMo-V2.6-Pro costs $0.3263/M input tokens and $0.6525/M output tokens, with separate rates for Cache Read at $0.003/M tokens.
MiMo-V2.6-Pro has a 1,050,000 token context window.
Yes. MiMo-V2.6-Pro accepts tools and tool_choice for function calling. It also supports structured outputs via a JSON schema in response_format.
MiMo-V2.6-Pro accepts text, images, video, and audio as input and returns text.
MiMo-V2.6-Pro is served by 5 providers on OpenRouter: GMICloud, DeepInfra, Xiaomi, StreamLake, and PrimaLabs. Requests are routed to the best available provider, with automatic failover to the others, and you can pin or exclude providers with provider routing.
MiMo-V2.6-Pro was released on September 21, 2026.