Moonshot AI model product
Kimi K2.5
5 is Moonshot AI's flagship agentic model and a new SOTA open model.
Updated Aug 12, 2026. Default version: Kimi K2.5
LLMBoard score66.0Kimi K2.5
Coverage80%38 benchmark families
Context window262.1KTokens
Official input price$0.60Moonshot AI API
Kimi K2.5 Specifications
Technical details for the model's default version.
- Version
- Kimi K2.5
- Released
- Jan 27, 2026
- Knowledge cutoff
- Jan 1, 2025
- Parameters
- 1T
- Context window
- 262.1K
- Max output
- 262.1K
- Inputs
- image, text, video
- Outputs
- text
- Open weights
- Yes
- License
- MIT
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.
| FinSearchComp T2&T3 | 67.8% | 01 | 1 | 100.0% | C | |
| InfoVQAtest | 92.6% | 01 | 12 | 100.0% | C | |
| MathVista-Mini | 90.1% | 01 | 23 | 100.0% | C | |
| MMVU | 80.4% | 01 | 4 | 100.0% | C | |
| OCRBench | 92.3% | 01 | 22 | 100.0% | C | |
| OJBench (C++) | 57.4% | 01 | 1 | 100.0% | C | |
| Seal-0 | 57.4% | 01 | 6 | 100.0% | C | |
| PaperBench | 63.5% | 02 | 3 | 50.0% | C | |
| WideSearch | 79.0% | 02 | 9 | 87.5% | C | |
| LongVideoBench | 79.8% | 03 | 4 | 33.3% | C | |
| MotionBench | 70.4% | 03 | 3 | 0.0% | C | |
| VideoMMMU | 86.6% | 03 | 26 | 92.0% | C | |
| AA-LCR | 70.0% | 04 | 16 | 80.0% | C | |
| ZEROBench | 0.11 points | 04 | 9 | 62.5% | C | |
| LVBench | 75.9% | 05 | 24 | 82.6% | C | |
| SimpleVQA | 0.712 points | 05 | 13 | 66.7% | C | |
| Video-MME | 87.4% | 05 | 17 | 75.0% | C | |
| WorldVQA | 46.3% | 05 | 5 | 0.0% | C | |
| HMMT 2025 | 95.4% | 06 | 33 | 84.4% | C | |
| LongBench v2 | 61.0% | 06 | 17 | 68.8% | C | |
| DeepSearchQA | 77.1% | 07 | 9 | 25.0% | C | |
| MMLU-Pro | 87.1% | 07 | 129 | 95.3% | C | |
| SciCode | 48.7% | 07 | 19 | 66.7% | C | |
| LiveCodeBench v6 | 85.0% | 09 | 53 | 84.6% | C | |
| OmniDocBench 1.5 | 88.8% | 09 | 17 | 50.0% | C | |
| MathVision | 84.2% | 10 | 32 | 71.0% | C | |
| CyberGym | 41.3% | 11 | 11 | 0.0% | C | |
| IMO-AnswerBench | 81.8% | 13 | 19 | 33.3% | C | |
| SWE-bench Multilingual | 73.0% | 13 | 34 | 63.6% | C | |
| FrontierSWE | 26.0% | 14 | 15 | 7.1% | B | |
| MMMU-Pro | 78.5% | 18 | 66 | 73.8% | C | |
| AIME 2025 | 96.1% | 19 | 114 | 84.1% | C | |
| Humanity's Last Exam | 50.2% | 20 | 93 | 79.3% | C | |
| BrowseComp | 74.9% | 26 | 58 | 56.1% | C | |
| CharXiv-R | 77.5% | 26 | 48 | 46.8% | C | |
| SWE-Bench Verified | 76.8% | 27 | 105 | 75.0% | C | |
| LiveBench | 69.1% | 32 | 38 | 16.2% | B | |
| Terminal-Bench 2.0 | 50.8% | 35 | 49 | 29.2% | C | |
| GPQA | 87.6% | 38 | 234 | 84.1% | C | |
| SWE-Bench Pro | 50.7% | 43 | 45 | 4.5% | C |
Kimi K2.5 Arena Results
Preference and agent-evaluation results for the default version.
| document | overall | 28 | 1429.5 | 19,891 | N/A | |
| document style control | overall | 31 | 1434.9 | 19,891 | N/A | |
| vision | overall | 33 | 1266.5 | 26,123 | N/A | |
| vision style control | overall | 41 | 1249.2 | 26,123 | N/A | |
| vision | overall | 43 | 1254.3 | 3,845 | N/A | |
| text | overall | 47 | 1445.4 | 70,609 | N/A | |
| webdev | overall | 49 | 1435.8 | 17,439 | N/A | |
| vision style control | overall | 51 | 1237.2 | 3,845 | N/A | |
| webdev | overall | 58 | 1405.2 | 3,109 | N/A | |
| text style control | overall | 60 | 1450.3 | 70,609 | N/A | |
| text factuality | overall | 65 | 1444.8 | 70,242 | N/A | |
| text style control | overall | 89 | 1430.7 | 8,194 | N/A | |
| text | overall | 94 | 1420.0 | 8,194 | N/A | |
| text factuality | overall | 98 | 1428.6 | 8,077 | N/A |
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
Browse runtime rankingsNo 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 Pricing
Official vendor API pricing appears first, followed by individual provider offers.
- Official API
- $0.60 input, $3 output per 1M
- Official provider
- Moonshot AI
- Lowest third-party
- From $0.30 input, $1.9 output per 1M via NanoGPT
- Tracked offerings
- 49
| Alibaba Coding Plan (China) | kimi-k2.5 | global | N/A | N/A | 262.1K | |
| Alibaba Coding Plan | kimi-k2.5 | global | N/A | N/A | 262.1K | |
| Alibaba Token Plan | kimi-k2.5 | global | N/A | N/A | 262.1K | |
| Tencent Coding Plan (China) | kimi-k2.5 | global | N/A | N/A | 262.1K | |
| Alibaba Token Plan (China) | kimi-k2.5 | global | N/A | N/A | 262.1K | |
| NanoGPT | moonshotai/kimi-k2.5 | global | $0.30 | $1.9 | 256K | |
| CrofAI | kimi-k2.5 | global | $0.35 | $1.7 | 262.1K | |
| DigitalOcean | kimi-k2.5 | global | $0.375 | $2.03 | 262.1K | |
| LLM Gateway | kimi-k2.5 | global | $0.405 | $1.98 | 262.1K | |
| SiliconFlow | moonshotai/Kimi-K2.5 | global | $0.45 | $2.25 | 262K | |
| Deep Infra | moonshotai/Kimi-K2.5 | global | $0.45 | $2.25 | 262.1K | |
| Meganova | moonshotai/Kimi-K2.5 | global | $0.45 | $2.8 | 262.1K | |
| Cortecs | kimi-k2.5 | global | $0.495 | $2.77 | 262.1K | |
| Nebius Token Factory | moonshotai/Kimi-K2.5 | global | $0.50 | $2.5 | 256K | |
| Together AI | moonshotai/Kimi-K2.5 | global | $0.50 | $2.8 | 262.1K | |
| Auriko | kimi-k2.5 | global | $0.50 | $2.8 | 262.1K | |
| TensorX | moonshotai/kimi-k2.5 | global | $0.50 | $2.8 | 262.1K | |
| Neuralwatt | moonshotai/Kimi-K2.5 | global | $0.52 | $2.59 | 262.1K | |
| Charm Hyper | kimi-k2.5 | global | $0.558 | $2.94 | 262.1K | |
| Venice AI | kimi-k2-5 | global | $0.56 | $3.5 | 256K | |
| OpenRouter | moonshotai/kimi-k2.5 | global | $0.57 | $2.85 | 262.1K | |
| Alibaba (China) | kimi-k2.5 | global | $0.574 | $2.41 | 262.1K | |
| ZenMux | moonshotai/kimi-k2.5 | global | $0.58 | $3.02 | 262K | |
| OpenCode Go | kimi-k2.5 | global | $0.60 | $3 | 262.1K | |
| Poe | novita/kimi-k2.5 | global | $0.60 | $3 | 128K | |
| Baseten | moonshotai/Kimi-K2.5 | global | $0.60 | $3 | 262K | |
| DaoXE | kimi-k2.5 | global | $0.60 | $3 | 262.1K | |
| OpenCode Zen | kimi-k2.5 | global | $0.60 | $3 | 262.1K | |
| FrogBot | kimi-k2.5 | global | $0.60 | $3 | 256K | |
| Abacus | kimi-k2.5 | global | $0.60 | $3 | 262.1K | |
| Azure Cognitive Services | kimi-k2.5 | global | $0.60 | $3 | 262.1K | |
| Moonshot AI (China) | kimi-k2.5 | global | $0.60 | $3 | 262.1K | |
| Azure | kimi-k2.5 | global | $0.60 | $3 | 262.1K | |
| AIHubMix | kimi-k2.5 | global | $0.60 | $3 | 262.1K | |
| Ofox | moonshotai/kimi-k2.5 | global | $0.60 | $3 | 262.1K | |
| OrcaRouter | kimi/kimi-k2.5 | global | $0.60 | $3 | 262.1K | |
| Amazon Bedrock | moonshotai.kimi-k2.5 | global | $0.60 | $3 | 262.1K | |
| Kilo Gateway | moonshotai/kimi-k2.5 | global | $0.60 | $3 | 262.1K | |
| Merge Gateway | moonshot/kimi-k2.5 | global | $0.60 | $3 | 262.1K | |
| Vercel AI Gateway | moonshotai/kimi-k2.5 | global | $0.60 | $3 | 262.1K | |
| NovitaAI | moonshotai/kimi-k2.5 | global | $0.60 | $3 | 262.1K | |
| Hugging Face | moonshotai/Kimi-K2.5 | global | $0.60 | $3 | 262.1K | |
| Cloudflare AI Gateway | workers-ai/@cf/moonshotai/kimi-k2.5 | global | $0.60 | $3 | 256K | |
| Jiekou.AI | moonshotai/kimi-k2.5 | global | $0.60 | $3 | 262.1K | |
| Moonshot AI | kimi-k2.5 | global | $0.60 | $3 | 262.1K | |
| HPC-AI | moonshotai/kimi-k2.5 | global | $0.60 | $3 | 256K | |
| CrossModel | moonshot/kimi-k2.5 | global | $0.62 | $3.3 | 262K |
Official prices use only the vendor's configured official Provider and positive standard USD PAYG rates. Third-party offers remain explicitly labeled.
Kimi K2.5 Versions
Available versions of this model. The score column identifies the version used in the overall ranking.
| Kimi K2.5 | 66.0 | 1T | 262.1K | 262.1K | Yes | MIT |
What is Kimi K2.5?
Key information about Kimi K2.5 and its available data.
5 is Moonshot AI's flagship agentic model and a new SOTA open model. It unifies vision and text, thinking and non-thinking modes, and single-agent and multi-agent execution into one model. Built with Full-Parameter RL tuning, it achieves state-of-the-art performance across agents, coding, image, and video benchmarks.
Data as of 2026-08-11.
Kimi K2.5 vs nearby models
Open a comparison with the three ranked models immediately above and below this model.
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
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?
Kimi K2.5's official API price is $0.60 per million input tokens and $3 per million output tokens via Moonshot AI. 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?
49 provider offerings are linked to the default version.
What models should I compare Kimi K2.5 with?
Nearby ranked alternatives include GLM 5, Kimi K2.7 Code, Qwen3.5 397B A17B.