moonshotai/kimi-k3.1-pro

Model release OpenRouter 模型 Source language: Chinese

OpenRouter 于 2026-10-24 收录该模型(收录时间不等于模型正式发布日期);上下文:1,048,576 tokens;输入价:$2.0000/百万 tokens;输出价:$10.0000/百万 tokens;定位(来源描述):Kimi K3.1 Pro is Moonshot AI's flagship open-weight multimodal reasoning model and a post-trained…

Full text

MoonshotAI: Kimi K3.1 Pro

moonshotai/kimi-k3.1-pro

Model weights

Compare

Kimi K3.1 Pro is Moonshot AI's flagship open-weight multimodal reasoning model and a post-trained successor to Kimi K3, with a 1M-token context window. It is built for long-horizon coding, 3D creation, financial research, and end-to-end agentic knowledge work, with native image and video understanding.

Modalities

In / out price

$0.2403 / $8.402per 1M

Context

1.0M

Released

Oct 23, 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

$0.2403$8.4023$0.2001.65s80 tok/s

99.94%

$0.2530$8.8000$0.1800.26s111 tok/s

100.00%

$0.7700$11.5570$0.2452.06s45 tok/s

100.00%

$1.0425$9.3500$0.1871.17s55 tok/s

99.83%

39% off

$2.00$1.2200$10.00$6.1000$0.20$0.1221.73s64 tok/s

99.97%

$1.4025$9.3500$0.1871.89s55 tok/s

99.85%

$1.7500$10.0000$0.1802.18s52 tok/s

99.91%

$1.8000$9.0000$0.1801.07s98 tok/s

100.00%

$1.8000$9.0000$0.1800.93s99 tok/s

99.97%

$1.8000$9.0000$0.1801.01s90 tok/s

100.00%

5% off

$2.00$1.9000$10.00$9.5000$0.20$0.1901.40s40 tok/s

99.89%

$2.0000$10.0000$0.2001.09s80 tok/s

98.44%

$2.0000$10.0000$0.2002.03s48 tok/s

99.56%

$2.0000$10.0000$0.2001.94s100 tok/s

99.72%

$2.0000$10.0000$0.2001.30s92 tok/s

99.97%

$2.0000$10.0000$0.2001.12s90 tok/s

100.00%

$2.0000$10.0000$0.2003.30s39 tok/s

97.59%

$2.0000$10.0000$0.2002.52s30 tok/s

100.00%

$2.3000$11.5000$0.2303.52s30 tok/s

99.66%

$2.7000$13.5000$0.2702.45s81 tok/s

99.70%

Fast

$2.7000$13.5000$0.2700.92s104 tok/s

100.00%

Fast

$2.7000$13.5000$0.2701.51s91 tok/s

99.97%

Highspeed

$4.0000$20.0000$0.2002.79s57 tok/s

99.79%

$2.0000$10.0000$0.2001.30s52 tok/s

83.51%

ProviderInput / Output /M

Input $0.2403 / Output $8.4023

80 tok/s99.94% uptime

Input $0.253 / Output $8.80

111 tok/s100.00% uptime

Input $0.77 / Output $11.557

45 tok/s100.00% uptime

Input $1.0425 / Output $9.35

55 tok/s99.83% uptime

Input $1.22 / Output $6.10

39% off

64 tok/s99.97% uptime

Input $1.4025 / Output $9.35

55 tok/s99.85% uptime

Input $1.75 / Output $10.00

52 tok/s99.91% uptime

Input $1.80 / Output $9.00

98 tok/s100.00% uptime

Input $1.80 / Output $9.00

99 tok/s99.97% uptime

Input $1.80 / Output $9.00

90 tok/s100.00% uptime

Input $1.90 / Output $9.50

5% off

40 tok/s99.89% uptime

Input $2.00 / Output $10.00

80 tok/s98.44% uptime

Input $2.00 / Output $10.00

48 tok/s99.56% uptime

Input $2.00 / Output $10.00

100 tok/s99.72% uptime

Input $2.00 / Output $10.00

92 tok/s99.97% uptime

Input $2.00 / Output $10.00

90 tok/s100.00% uptime

Input $2.00 / Output $10.00

39 tok/s97.59% uptime

Input $2.00 / Output $10.00

30 tok/s100.00% uptime

Input $2.30 / Output $11.50

30 tok/s99.66% uptime

Input $2.70 / Output $13.50

81 tok/s99.70% uptime

Input $2.70 / Output $13.50

Fast

104 tok/s100.00% uptime

Input $2.70 / Output $13.50

Fast

91 tok/s99.97% uptime

Input $4.00 / Output $20.00

Highspeed

57 tok/s99.79% uptime

Input $2.00 / Output $10.00

52 tok/s83.51% 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

$0.3085/M tokens

Output

$10/M tokens

Kimi K3.1 Pro — Price history

Price history

No effective price data for this selection

$0.184$8.8094.2%8.4%

$0.202$8.4194.1%4.7%

$0.210$6.1092.0%2.0%

$0.233$9.3594.6%0.3%

$0.242$9.0991.9%3.1%

$0.250$9.0095.7%6.5%

$0.267$10.0094.5%1.9%

$0.274$9.0094.2%2.3%

$0.276$9.0094.1%4.3%

$0.280$10.0095.6%15.9%

$0.282$10.0095.5%9.9%

$0.300$10.0094.5%6.4%

$0.303$10.0094.3%3.0%

$0.341$10.0092.2%0.7%

Fast

$0.374$13.5095.7%2.3%

$0.380$10.0090.0%2.3%

Fast

$0.393$13.5095.0%1.2%

$0.417$13.5094.0%1.2%

$0.422$11.5090.7%21.9%

$0.442$9.3579.1%0.0%

$0.506$9.5181.6%0.1%

$0.508$10.0082.9%0.1%

$0.522$10.0082.1%0.8%

Highspeed

$0.633$20.0088.6%0.5%

ProviderInput / Output /M

Token share

Effective in $0.184 / Effective out $8.80

Token share

8.4%

Effective in $0.202 / Effective out $8.41

Token share

4.7%

Effective in $0.210 / Effective out $6.10

Token share

2.0%

Effective in $0.233 / Effective out $9.35

Token share

0.3%

Effective in $0.242 / Effective out $9.09

Token share

3.1%

Effective in $0.250 / Effective out $9.00

Token share

6.5%

Effective in $0.267 / Effective out $10.00

Token share

1.9%

Effective in $0.274 / Effective out $9.00

Token share

2.3%

Effective in $0.276 / Effective out $9.00

Token share

4.3%

Effective in $0.280 / Effective out $10.00

Token share

15.9%

Effective in $0.282 / Effective out $10.00

Token share

9.9%

Effective in $0.300 / Effective out $10.00

Token share

6.4%

Effective in $0.303 / Effective out $10.00

Token share

3.0%

Effective in $0.341 / Effective out $10.00

Token share

0.7%

Effective in $0.374 / Effective out $13.50

FastToken share

2.3%

Effective in $0.380 / Effective out $10.00

Token share

2.3%

Effective in $0.393 / Effective out $13.50

FastToken share

1.2%

Effective in $0.417 / Effective out $13.50

Token share

1.2%

Effective in $0.422 / Effective out $11.50

Token share

21.9%

Effective in $0.442 / Effective out $9.35

Token share

0.0%

Effective in $0.506 / Effective out $9.51

Token share

0.1%

Effective in $0.508 / Effective out $10.00

Token share

0.1%

Effective in $0.522 / Effective out $10.00

Token share

0.8%

Effective in $0.633 / Effective out $20.00

HighspeedToken share

0.5%

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

111tok/s

P50, best across providers

Latency

0.26s

P50, best provider

Reasoning effortLatency / throughput

AutoExacto benchmarks

GPQA DiamondTAU-BenchMoonshot AI (Highspeed)93.3%78.7%DigitalOcean93.3%78.7%Phala92.9%78.7%Fireworks92.8%78.7%Nebius Token Factory93.9%77.3%Moonshot AI92.6%75.3%Phala92.4%75.0%

+14 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.

99.90%

Availability over the last 3 days

Last 72 hours

Availability 99.90%

3 Days Ago2 Days AgoYesterdayNow

Availability over the last 24 hours

OpenRouter Availability

99.83%

Without Routing

97.14%

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 MoonshotAI: Kimi K3.1 Pro (Artificial Analysis, Design Arena and Vals AI)

SourceBenchmarkScore

Artificial AnalysisKimi K3.1 Pro Intelligence Index46.5

Artificial AnalysisKimi K3.1 Pro HLE52.5%

Artificial AnalysisKimi K3.1 Pro AA-LCR86.3%

Artificial AnalysisKimi K3.1 Pro GDPval-AA68.0%

Artificial AnalysisKimi K3.1 Pro CritPt24.6%

Artificial AnalysisKimi K3.1 Pro SciCode60.8%

Artificial AnalysisKimi K3.1 Pro Terminal-Bench 4.067.2%

Artificial AnalysisKimi K3.1 Pro AA-Omniscience Accuracy48.9%

Artificial AnalysisKimi K3.1 Pro AA-Omniscience Non-Hallucination Rate62.7%

Design ArenaKimi K3.1 Pro Agents Arena Androidnative Elo1153

Design ArenaKimi K3.1 Pro Agents Arena Codetovideo Elo1235

Design ArenaKimi K3.1 Pro Agents Arena Full Stack Elo1321

Design ArenaKimi K3.1 Pro Agents Arena Mobile Apps Elo1267

Design ArenaKimi K3.1 Pro Agents Arena Pdf Elo1217

Design ArenaKimi K3.1 Pro Agents Arena Python-Pptxslides Elo1279

Design ArenaKimi K3.1 Pro Agents Arena Spreadsheet Elo1326

Design ArenaKimi K3.1 Pro Agents Arena Webapps Elo1312

Design ArenaKimi K3.1 Pro Models Arena 3D Elo1443

Design ArenaKimi K3.1 Pro Models Arena Asciiart Elo1294

Design ArenaKimi K3.1 Pro Models Arena Code Categories Elo1350

Design ArenaKimi K3.1 Pro Models Arena Data Visualization Elo1297

Design ArenaKimi K3.1 Pro Models Arena Game Development Elo1403

Design ArenaKimi K3.1 Pro Models Arena SVG Elo1329

Design ArenaKimi K3.1 Pro Models Arena UI Component Elo1389

Design ArenaKimi K3.1 Pro Models Arena Website Elo1321

Vals AIKimi K3.1 Pro Access to Justice Benchmark69.5%

Vals AIKimi K3.1 Pro BioMysteryBench75.6%

Vals AIKimi K3.1 Pro CyberOffense53.8%

Vals AIKimi K3.1 Pro EMB v1.170.0%

Vals AIKimi K3.1 Pro Finance Agent v2.161.5%

Vals AIKimi K3.1 Pro Vals Finance Index56.0%

Vals AIKimi K3.1 Pro HLAB-Vals-1.111.7%

Vals AIKimi K3.1 Pro IOI81.9%

Vals AIKimi K3.1 Pro Vals Legal Index67.0%

Vals AIKimi K3.1 Pro Legal Research Bench37.5%

Vals AIKimi K3.1 Pro MysteryMechanism26.1%

Vals AIKimi K3.1 Pro ProofBench v1.191.0%

Vals AIKimi K3.1 Pro Vals Science Index44.4%

Vals AIKimi K3.1 Pro SRE Bench27.1%

Vals AIKimi K3.1 Pro Tax Agent Bench24.9%

Vals AIKimi K3.1 Pro Terminal-Bench 4.063.6%

Vals AIKimi K3.1 Pro Terminal-Bench-Science31.4%

Vals AIKimi K3.1 Pro Vals Index62.5%

Vals AIKimi K3.1 Pro Vibe Code Bench v1.184.9%

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.

pi

There are many coding agents, but this one is yours.

273Btokens

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.

226Btokens

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.

208Btokens

4.

omp

new

140Btokens

5.

Codex

A coding agent that helps you build and ship with AI

129Btokens

Rest of OpenRouter

215B

Top 5 Apps

31.5B

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

Frequently asked questions

Kimi K3.1 Pro is Moonshot AI's flagship open-weight multimodal reasoning model and a post-trained successor to Kimi K3, with a 1M-token context window. It is built for long-horizon coding, 3D creation, financial research, and end-to-end agentic knowledge work, with native image and video understanding.

Kimi K3.1 Pro costs $0.2403/M input tokens and $8.402/M output tokens, with separate rates for Cache Read at $0.20/M tokens.

Kimi K3.1 Pro has a 1,048,576 token context window. It supports up to 1,048,576 completion tokens.

Yes. Kimi K3.1 Pro accepts tools and tool_choice for function calling. It also supports structured outputs via a JSON schema in response_format.

Kimi K3.1 Pro accepts text, images, and video as input and returns text.

Kimi K3.1 Pro is served by 18 providers on OpenRouter: Relace, Wafer, Morph, Sail Research, Decart, AkashML, inference.net, DeepInfra and 10 more. Requests are routed to the best available provider, with automatic failover to the others, and you can pin or exclude providers with provider routing.

Kimi K3.1 Pro was released on October 23, 2026.

Sources

← Back to the daily brief