Skip to content

Moonshot AI · Chat

Kimi K2 Thinking

Moonshot AI 🇨🇳ChatKimi familyOpen weights. One of its sources says the weights are publicly available.The maker lists this model as deprecated (29 Mar 2026).
From Neosantara's listing:
Kimi K2 Thinking is Moonshot AI's reasoning model with chain-of-thought capabilities for complex problem solving in math, coding, and logic.
llminfo IQ
123.8
Max context
262K
Listings
38

Evidence and sources

International API storefront is Singapore-based; principal operating entity 北京月之暗面科技有限公司 (Beijing).

🤗 moonshotai/Kimi-K2-Thinkingllminfo IQ 123.8, ranked among chat models. From 184 benchmarks. Plenty of data behind it.How llminfo IQ is computedCoding IQ 100.2, 95% interval 94–106. From 10 benchmark families across 8 sources. Plenty of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ.How Coding IQ is computedPoor value (41.6). Costs more per task than its benchmark results predict. From 33 benchmarks with measured task costs. Plenty of data behind it.How Value for money worksYap score: tokens used per coding task vs. models with the same resultsTypical (50 ±18). Generates about as many tokens as models with similar coding results. Measured at default effort. Little data behind it, so treat it as a rough estimate.Median 24,809 output tokens per task at default effort.How the Yap score worksReleased 6 Nov 2025Maker HQ: CN (source)
All rankings: #228 on NanoGPT recently #316 on OpenRouter this month #334 on OpenRouter this week #338 on OpenRouter todaySources: GMI Cloud data, PPIO data, LiteLLM data, Models.dev data, AI/ML API data, Neosantara Inference data, llminfo researchPrimary data from GMI Cloud from PPIO: name from LiteLLM: deprecation date from Models.dev: weights availability from AI/ML API: primary mode from Neosantara Inference: description from llminfo research: release date

Advantages and trade-offs

Compared with 10 chat models, picked mainly by popularity and IQ

Advantages

Open weights available

The weights are public. 2 of the 9 peers that say either way have open weights.

Trade-offs

Deprecated by its maker

The maker has already deprecated this model (29 Mar 2026).

We only make a claim when there's enough data on models of the same type. Updated 22 Nov, 06:07 UTC.Prices come from the model's maker where they sell it, otherwise their official partner, otherwise the market median. This comparison uses 10 maker's prices.

Input & output

as stated by the model's providers
Text
Text

Capabilities

Supported operations and provider-stated features

ReasoningFunction callingStructured outputPrompt cachingWeb searchBatch API

Listings

 Moonshot AIkimi-k2-thinking-turboSource: GenAI Prices data262K$1.15cache read $0.15$8.00
 Tokenmixkimi-k2-thinkingSources: TokenMix data, TokenMix PricingNot published by the provider$0.5294$2.1176
 Aihubmixkimi-k2-thinkingSource: AIHubMix data docs262K$0.548cache read $0.137$2.192
 Alibaba Cloud (China)kimi-k2-thinkingSource: Models.dev data262K$0.574$2.294
 Ppiomoonshotai/kimi-k2-thinkingSource: PPIO data262K
 Qiniumoonshotai/kimi-k2-thinkingSources: Models.dev data, Qiniu AI data, Qiniu AI Model Plaza Pricing256K
 Amazon Web Servicesbedrock_mantle/moonshotai.kimi-k2-thinkingSource: LiteLLM data256K$0.60batch $0.30$2.50batch $1.25
 Microsoft Azureazure_ai/kimi-k2-thinkingSource: LiteLLM data262K$0.60$2.50
 Googlevertex_ai/moonshotai/kimi-k2-thinking-maas · maasSources: LiteLLM data, Models.dev data256K$0.60cache read $0.06$2.50
 Flowdotmoonshotai/kimi-k2-thinkingSource: FlowDot data262K$0.60$2.50
Show all 33 listings

Showing 10 of 33: the maker's own listing first, then the cheapest per 1M input tokens from each provider.

Also sold by 302ai, Aimlapi, Baseten, Crusoe, Evolink, Fireworks AI, Gmicloud, Hugging Face, Ionet, Kilo, Meganova, NanoGPT, Neosantara, Nous Portal, Novita AI, OpenRouter, Replicate, Tencent Lkeap.

5 listings without a price are hidden.

Benchmarks

Named effort levels run from lowest to highest. Select a result to see its source details.

Coding

SciCodeSource: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim
SWE-bench Verified5 related benchmarks · 5 results
SWE-bench Verified
Source: Hugging Face Official Benchmarks data · Reported by the model's developer. The source relaying it did not verify it. claim
SWE-bench Multilingual
Source: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim

Math

HMMT 2025Source: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim

Reasoning

CL-benchSource: Epoch AI Benchmarking Hub data
FRAMESSource: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim
OJBenchSource: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim
Seal-0Source: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim

Agentic

OpenHandsSource: OpenHands Index data

Science

GPQASource: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim

Vision

Factuality

Knowledge

MMLU-ProSource: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim

General

Human preference

Multilingual

Other

Nejumi 428 related benchmarks · 28 results
ValsSource: Vals AI

Data credits: Data by LLM Stats · Epoch AI Benchmarking Hub, CC BY 4.0 (licence) · EQ-Bench (Samuel J. Paech) — cite paech2023eqbench · OpenEvals/leaderboard-data, MIT (licence) · HalluHard: Fan, Delsad, Flammarion and Andriushchenko (EPFL), arXiv:2602.01031 · MathArena (ETH SRI / INSAIT), MIT — cite dekoninck2026matharena (licence) · Nejumi Leaderboard 4, Weights & Biases Japan; public publisher-selected numeric evaluation results · OpenHands Index (All Hands AI), Apache-2.0 (licence) · SnakeBench (Greg Kamradt / ARC Prize Foundation), MIT (licence) · Hugging Face official benchmark leaderboards; licence varies by benchmark

Also known as: accounts/fireworks/models/kimi-k2-thinking, alicloud-kimi-k2-thinking, azure_ai/kimi-k2-thinking, baseten/moonshotai/kimi-k2-thinking, bedrock/ap-northeast-1/moonshotai.kimi-k2-thinking, bedrock/ap-south-1/moonshotai.kimi-k2-thinking, bedrock/moonshotai.kimi-k2-thinking, bedrock/sa-east-1/moonshotai.kimi-k2-thinking, bedrock/us-east-1/moonshotai.kimi-k2-thinking, bedrock/us-east-2/moonshotai.kimi-k2-thinking, bedrock/us-west-2/moonshotai.kimi-k2-thinking, bedrock_mantle/ap-northeast-1/moonshotai.kimi-k2-thinking, bedrock_mantle/ap-south-1/moonshotai.kimi-k2-thinking, bedrock_mantle/ap-southeast-2/moonshotai.kimi-k2-thinking, bedrock_mantle/ap-southeast-3/moonshotai.kimi-k2-thinking, bedrock_mantle/eu-north-1/moonshotai.kimi-k2-thinking, bedrock_mantle/eu-west-2/moonshotai.kimi-k2-thinking, bedrock_mantle/moonshotai.kimi-k2-thinking, bedrock_mantle/sa-east-1/moonshotai.kimi-k2-thinking, cc-kimi-k2-thinking, crusoe/moonshotai/kimi-k2-thinking, fireworks_ai/accounts/fireworks/models/kimi-k2-thinking, gmi/moonshotai/kimi-k2-thinking, kimi-k2-thinking-251104, kimi-k2-thinking-turbo, moonshot ai/kimi-k2-thinking, moonshot/kimi-k2-thinking, moonshotai/kimi-k2-thinking, moonshotai/kimi-k2-thinking-maas, novita/kimi-k2-thinking, novita/moonshotai/kimi-k2-thinking, pro/moonshotai/kimi-k2-thinking, replicate/moonshotai/kimi-k2-thinking, sophnet/kimi-k2-thinking, vertex_ai/moonshotai/kimi-k2-thinking-maas

Like this one. Tap or hover any dotted value, label or icon to see where it comes from and how it is measured. open details when tapped or hovered. How this data works