Moonshot AI model product
Kimi K2.5
5 is an open large language model from Moonshot AI that combines vision and text with thinking and non-thinking modes.
Updated Sep 8, 2026. Default version: Kimi K2.5
LLMBoard Score65.6Kimi K2.5
Coverage60%38 benchmark families
Context window262.1KTokens
Official input priceN/AOfficial price unavailable
Kimi K2.5 Capability Profile
This profile uses the model's current scored version. Arena ratings and prices are shown separately.
Kimi K2.5 Benchmark Results
Benchmark scores for Kimi K2.5.
Sort by
Benchmark | Score | Rank | Participants | Percentile | Evidence | Evaluated |
|---|
| BenchmarkFinSearchComp T2&T3 | Score67.80% | Rank01 | Participants1 | Percentile100.00% | EvidenceC | Evaluated |
| BenchmarkInfoVQAtest | Score92.60% | Rank01 | Participants12 | Percentile100.00% | EvidenceC | Evaluated |
| BenchmarkMathVista-Mini | Score90.10% | Rank01 | Participants25 | Percentile100.00% | EvidenceC | Evaluated |
| BenchmarkOCRBench | Score92.30% | Rank01 | Participants24 | Percentile100.00% | EvidenceC | Evaluated |
| BenchmarkOJBench (C++) | Score57.40% | Rank01 | Participants1 | Percentile100.00% | EvidenceC | Evaluated |
| BenchmarkSeal-0 | Score57.40% | Rank01 | Participants6 | Percentile100.00% | EvidenceC | Evaluated |
| BenchmarkMMVU | Score80.40% | Rank02 | Participants5 | Percentile75.00% | EvidenceC | Evaluated |
| BenchmarkPaperBench | Score63.50% | Rank02 | Participants3 | Percentile50.00% | EvidenceC | Evaluated |
| BenchmarkLongVideoBench | Score79.80% | Rank03 | Participants4 | Percentile33.33% | EvidenceC | Evaluated |
| BenchmarkMotionBench | Score70.40% | Rank03 | Participants3 | Percentile0.00% | EvidenceC | Evaluated |
| BenchmarkVideoMMMU | Score86.60% | Rank03 | Participants26 | Percentile92.00% | EvidenceC | Evaluated |
| BenchmarkWideSearch | Score79.00% | Rank04 | Participants11 | Percentile70.00% | EvidenceC | Evaluated |
| BenchmarkAA-LCR | Score70.00% | Rank05 | Participants19 | Percentile77.78% | EvidenceC | Evaluated |
| BenchmarkSimpleVQA | Score0.71 points | Rank05 | Participants14 | Percentile69.23% | EvidenceC | Evaluated |
| BenchmarkVideo-MME | Score87.40% | Rank05 | Participants17 | Percentile75.00% | EvidenceC | Evaluated |
| BenchmarkWorldVQA | Score46.30% | Rank05 | Participants5 | Percentile0.00% | EvidenceC | Evaluated |
| BenchmarkZEROBench | Score0.11 points | Rank05 | Participants10 | Percentile55.56% | EvidenceC | Evaluated |
| BenchmarkHMMT 2025 | Score95.40% | Rank06 | Participants33 | Percentile84.38% | EvidenceC | Evaluated |
| BenchmarkLongBench v2 | Score61.00% | Rank06 | Participants17 | Percentile68.75% | EvidenceC | Evaluated |
| BenchmarkDeepSearchQA | Score77.10% | Rank08 | Participants10 | Percentile22.22% | EvidenceC | Evaluated |
| BenchmarkMMLU-Pro | Score87.10% | Rank08 | Participants138 | Percentile94.89% | EvidenceC | Evaluated |
| BenchmarkSciCode | Score48.70% | Rank08 | Participants24 | Percentile69.57% | EvidenceC | Evaluated |
| BenchmarkLVBench | Score75.90% | Rank09 | Participants28 | Percentile70.37% | EvidenceC | Evaluated |
| BenchmarkOmniDocBench 1.5 | Score88.80% | Rank10 | Participants18 | Percentile47.06% | EvidenceC | Evaluated |
| BenchmarkLiveCodeBench v6 | Score85.00% | Rank13 | Participants62 | Percentile80.33% | EvidenceC | Evaluated |
| BenchmarkMathVision | Score84.20% | Rank13 | Participants35 | Percentile64.71% | EvidenceC | Evaluated |
| BenchmarkIMO-AnswerBench | Score81.80% | Rank14 | Participants20 | Percentile31.58% | EvidenceC | Evaluated |
| BenchmarkCyberGym | Score41.30% | Rank15 | Participants15 | Percentile0.00% | EvidenceC | Evaluated |
| BenchmarkFrontierSWE | Score26.00% | Rank15 | Participants16 | Percentile6.67% | EvidenceB | Evaluated |
| BenchmarkSWE-bench Multilingual | Score73.00% | Rank17 | Participants43 | Percentile61.90% | EvidenceC | Evaluated |
Kimi K2.5 Arena Results
Preference and agent-evaluation results for the default version.
Sort by
Arena | Category | Rank | Rating / score | Votes | Observations | Result date |
|---|
| Arenavision | Categoryentity recognition | Rank09 | Rating / score1,279.65 | Votes242 | ObservationsN/A | Result date |
| Arenavision | Categorycaptioning | Rank11 | Rating / score1,254.46 | Votes108 | ObservationsN/A | Result date |
| Arenavision style control | Categorycaptioning | Rank13 | Rating / score1,198.07 | Votes108 | ObservationsN/A | Result date |
| Arenavision style control | Categoryentity recognition | Rank13 | Rating / score1,228.13 | Votes242 | ObservationsN/A | Result date |
| Arenatext factuality | Categorykorean | Rank15 | Rating / score1,414.02 | Votes739 | ObservationsN/A | Result date |
| Arenavision | Categoryhumor | Rank15 | Rating / score1,284.58 | Votes978 | ObservationsN/A | Result date |
| Arenavision style control | Categoryhumor | Rank19 | Rating / score1,263.50 | Votes978 | ObservationsN/A | Result date |
| Arenatext factuality | Categorygerman | Rank20 | Rating / score1,416.46 | Votes716 | ObservationsN/A | Result date |
| Arenatext factuality | Categoryjapanese | Rank20 | Rating / score1,416.59 | Votes503 | ObservationsN/A | Result date |
| Arenavision | Categoryhomework | Rank22 | Rating / score1,296.41 | Votes3,512 | ObservationsN/A | Result date |
| Arenavision style control | Categorychinese | Rank23 | Rating / score1,303.33 | Votes236 | ObservationsN/A | Result date |
| Arenadocument | Categoryoverall | Rank24 | Rating / score1,429.49 | Votes19,891 | ObservationsN/A | Result date |
| Arenatext | Categorymath | Rank25 | Rating / score1,469.77 | Votes3,925 | ObservationsN/A | Result date |
| Arenatext factuality | Categorycoding | Rank25 | Rating / score1,515.71 | Votes937 | ObservationsN/A | Result date |
| Arenavision | Categorychinese | Rank25 | Rating / score1,320.43 | Votes236 | ObservationsN/A | Result date |
| Arenavision | Categorycreative writing vision | Rank25 | Rating / score1,266.31 | Votes1,621 | ObservationsN/A | Result date |
| Arenavision style control | Categoryhomework | Rank25 | Rating / score1,289.50 | Votes3,512 | ObservationsN/A | Result date |
| Arenatext factuality | Categorychinese | Rank26 | Rating / score1,505.06 | Votes2,953 | ObservationsN/A | Result date |
| Arenatext factuality | Categoryspanish | Rank26 | Rating / score1,447.77 | Votes1,802 | ObservationsN/A | Result date |
| Arenadocument style control | Categoryoverall | Rank27 | Rating / score1,434.87 | Votes19,891 | ObservationsN/A | Result date |
| Arenatext factuality | Categoryindustry mathematical | Rank28 | Rating / score1,459.23 | Votes2,904 | ObservationsN/A | Result date |
| Arenatext factuality | Categorypolish | Rank28 | Rating / score1,423.14 | Votes947 | ObservationsN/A | Result date |
| Arenavision | Categoryenglish | Rank28 | Rating / score1,272.12 | Votes10,398 | ObservationsN/A | Result date |
| Arenatext factuality | Categorymath | Rank29 | Rating / score1,456.62 | Votes2,980 | ObservationsN/A | Result date |
| Arenawebdev | Categoryimage to webdev | Rank29 | Rating / score1,439.81 | Votes1,532 | ObservationsN/A | Result date |
| Arenatext | Categorycoding | Rank30 | Rating / score1,484.89 | Votes1,776 | ObservationsN/A | Result date |
| Arenatext | Categoryindustry mathematical | Rank30 | Rating / score1,470.82 | Votes3,724 | ObservationsN/A | Result date |
| Arenavision | Categoryocr | Rank30 | Rating / score1,273.59 | Votes18,469 | ObservationsN/A | Result date |
| Arenavision | Categoryoverall | Rank30 | Rating / score1,266.74 | Votes26,728 | ObservationsN/A | Result date |
| Arenatext | Categoryjapanese | Rank31 | Rating / score1,418.89 | Votes808 | ObservationsN/A | Result date |
Kimi K2.5 Pricing
Official vendor API pricing appears first, followed by individual provider offers.
- Official API
- N/A
- Official provider
- N/A
- Lowest third-party
- From $0.30 input, $1.9 output per 1M via NanoGPT
- Tracked offerings
- 45
Sort by
Provider | Provider model ID | Region | Input / 1M | Output / 1M | Context | Updated |
|---|
| ProviderAlibaba Token Plan | Provider model IDkimi-k2.5 | Regionglobal | Input / 1MN/A | Output / 1MN/A | Context262.1K | Updated |
| ProviderAlibaba Coding Plan (China) | Provider model IDkimi-k2.5 | Regionglobal | Input / 1MN/A | Output / 1MN/A | Context262.1K | Updated |
| ProviderAlibaba Coding Plan | Provider model IDkimi-k2.5 | Regionglobal | Input / 1MN/A | Output / 1MN/A | Context262.1K | Updated |
| ProviderTencent Coding Plan (China) | Provider model IDkimi-k2.5 | Regionglobal | Input / 1MN/A | Output / 1MN/A | Context262.1K | Updated |
| ProviderSCNet Token Plan | Provider model IDKimi-K2.5 | Regionglobal | Input / 1MN/A | Output / 1MN/A | Context262.1K | Updated |
| ProviderAlibaba Token Plan (China) | Provider model IDkimi-k2.5 | Regionglobal | Input / 1MN/A | Output / 1MN/A | Context262.1K | Updated |
| ProviderNanoGPT | Provider model IDmoonshotai/kimi-k2.5 | Regionglobal | Input / 1M$0.30 | Output / 1M$1.9 | Context256K | Updated |
| ProviderDevPass (LLM Gateway) | Provider model IDkimi-k2.5 | Regionglobal | Input / 1M$0.405 | Output / 1M$1.98 | Context262.1K | Updated |
| ProviderMeganova | Provider model IDmoonshotai/Kimi-K2.5 | Regionglobal | Input / 1M$0.45 | Output / 1M$2.8 | Context262.1K | Updated |
| ProviderDeep Infra | Provider model IDmoonshotai/Kimi-K2.5 | Regionglobal | Input / 1M$0.45 | Output / 1M$2.25 | Context262.1K | Updated |
| ProviderOpenRouter | Provider model IDmoonshotai/kimi-k2.5 | Regionglobal | Input / 1M$0.45 | Output / 1M$2.25 | Context262.1K | Updated |
| ProviderSiliconFlow | Provider model IDmoonshotai/Kimi-K2.5 | Regionglobal | Input / 1M$0.45 | Output / 1M$2.25 | Context262K | Updated |
| ProviderCortecs | Provider model IDkimi-k2.5 | Regionglobal | Input / 1M$0.495 | Output / 1M$2.77 | Context262.1K | Updated |
| ProviderAuriko | Provider model IDkimi-k2.5 | Regionglobal | Input / 1M$0.50 | Output / 1M$2.8 | Context262.1K | Updated |
| ProviderDigitalOcean | Provider model IDkimi-k2.5 | Regionglobal | Input / 1M$0.50 | Output / 1M$2.7 | Context262.1K | Updated |
| ProviderTogether AI | Provider model IDmoonshotai/Kimi-K2.5 | Regionglobal | Input / 1M$0.50 | Output / 1M$2.8 | Context262.1K | Updated |
| ProviderTensorX | Provider model IDmoonshotai/kimi-k2.5 | Regionglobal | Input / 1M$0.50 | Output / 1M$2.8 | Context262.1K | Updated |
| ProviderCharm Hyper | Provider model IDkimi-k2.5 | Regionglobal | Input / 1M$0.5584 | Output / 1M$2.94 | Context262.1K | Updated |
| ProviderVenice AI | Provider model IDkimi-k2-5 | Regionglobal | Input / 1M$0.56 | Output / 1M$3.5 | Context256K | Updated |
| ProviderAlibaba (China) | Provider model IDkimi-k2.5 | Regionglobal | Input / 1M$0.574 | Output / 1M$2.41 | Context262.1K | Updated |
| ProviderZenMux | Provider model IDmoonshotai/kimi-k2.5 | Regionglobal | Input / 1M$0.58 | Output / 1M$3.02 | Context262K | Updated |
| ProviderHPC-AI | Provider model IDmoonshotai/kimi-k2.5 | Regionglobal | Input / 1M$0.60 | Output / 1M$3 | Context256K | Updated |
| ProviderOrcaRouter | Provider model IDkimi/kimi-k2.5 | Regionglobal | Input / 1M$0.60 | Output / 1M$3 | Context262.1K | Updated |
| ProviderJalapeno Cloud | Provider model IDKimi-K2.5 | Regionglobal | Input / 1M$0.60 | Output / 1M$3 | Context262.1K | Updated |
| ProviderHugging Face | Provider model IDmoonshotai/Kimi-K2.5 | Regionglobal | Input / 1M$0.60 | Output / 1M$3 | Context262.1K | Updated |
| ProviderJiekou.AI | Provider model IDmoonshotai/kimi-k2.5 | Regionglobal | Input / 1M$0.60 | Output / 1M$3 | Context262.1K | Updated |
| ProviderDaoXE | Provider model IDkimi-k2.5 | Regionglobal | Input / 1M$0.60 | Output / 1M$3 | Context262.1K | Updated |
| ProviderAmazon Bedrock | Provider model IDmoonshotai.kimi-k2.5 | Regionglobal | Input / 1M$0.60 | Output / 1M$3 | Context262.1K | Updated |
| ProviderOpenCode Go | Provider model IDkimi-k2.5 | Regionglobal | Input / 1M$0.60 | Output / 1M$3 | Context262.1K | Updated |
| ProviderMerge Gateway | Provider model IDmoonshot/kimi-k2.5 | Regionglobal | Input / 1M$0.60 | Output / 1M$3 | Context262.1K | Updated |
Kimi K2.5 Runtime Performance
Provider-specific output speed and catalog latency for Kimi K2.5. Runtime does not affect the capability score.
No runtime data
No provider-specific speed or latency record is linked to the default version yet.
Output Speed is generated output tokens received per second. Catalog latency is reported separately from observed provider TTFT.
Kimi K2.5 Specifications
Technical details for the model's default version.
- Version
- Kimi K2.5
- Released
- Jan 27, 2026
- Knowledge cutoff
- Unknown
- Parameters
- 1T
- Context window
- 262.1K
- Max output
- 262.1K
- Inputs
- image, text, video
- Outputs
- text
- Open weights
- Yes
- License
- MIT
Kimi K2.5 Versions
Available versions of this model. The score column identifies the version used in the overall ranking.
Sort by
Version | Released | LLMBoard | Parameters | Context | Max output | Open weights | License |
|---|
| VersionKimi K2.5 | Released | LLMBoard65.57 | Parameters1T | Context262.1K | Max output262.1K | Open weightsYes | LicenseMIT |
Models similar to Kimi K2.5
Recommendations prioritize the same model type and family, then the closest LLMBoard score.
MA
Kimi K2.7 Code
Moonshot AI
MA
Kimi K2 Thinking
Moonshot AI
MA
Kimi K2.6
Moonshot AI
MA
Kimi K3
Moonshot AI
MA
Kimi K2
Moonshot AI
MA
Kimi k1.5
Moonshot AI
What is Kimi K2.5?
Key information about Kimi K2.5 and its available data.
5 is Moonshot AI's flagship agentic model, supporting vision and text, thinking and non-thinking modes, and single-agent and multi-agent execution. It uses Full-Parameter RL tuning and is evaluated across agent, coding, image, and video benchmarks.
Data as of 2026-09-08.
FAQ
Common questions about Kimi K2.5.
When was Kimi K2.5 released?
Kimi K2.5's default version was released on Jan 27, 2026.
How much does Kimi K2.5 cost?
No official standard PAYG price is currently available for Kimi K2.5. The lowest tracked third-party offer starts at $0.30 input and $1.9 output via NanoGPT.
Who created Kimi K2.5?
Kimi K2.5 was created by Moonshot AI.
What is the context window for Kimi K2.5?
The default version has a 262.1K token context window.
Is Kimi K2.5 open weight?
Yes. The default version is marked as open weight under MIT.
How many API providers offer Kimi K2.5?
45 provider offerings are linked to the default version.