upcube
Kimi•Open weights•January 27, 2026

Kimi K2.5

Intelligence, Performance & Price Analysis

Source:Artificial Analysis scrapeType:Open weights
Intelligence
38

score

Kimi K2.5 (Reasoning) scores 38 (estimated) on the Artificial Analysis Intelligence Index, placing …

vs class avg: 25

Speed
43.1

output tokens/sec

Kimi K2.5 (Reasoning) generates output at 43.1 tokens per second (based on Kimi's API), which is at …

Latency
3.00s

TTFT

Kimi K2.5 (Reasoning) has a time to first token (TTFT) of 3.00s (based on Kimi's API), which is at …

Input Price
$0.60

/ 1M tokens

Kimi K2.5 (Reasoning) costs $0.60 per 1M input tokens (somewhat higher than average, median: $0. …

Output Price
$3.00

/ 1M tokens

Kimi K2.5 (Reasoning) costs $0.60 per 1M input tokens (somewhat higher than average, median: $0. …

Decision Overview

Ethen's intelligence-driven routing assessment for this model

Overall verdict

Frontier research / complex reasoning

Reserve this model for the hardest requests; simpler or repetitive work should stay on cheaper routes.

Best for

  • Complex reasoning
  • Research synthesis
  • Hard coding tasks
  • Multi-step agentic workflows

Weak or unsuitable

  • High-volume chat
  • Simple classification
  • Cost-sensitive pipelines
  • Predictable high-volume throughput

Routing recommendation

1

Cheaper routes for predictable extraction, labeling, or summarization

2

Cache-backed reuse for repeated prompts

Cost pressure

Moderate pricing — $0.60 / M input, $3.00 / M output. Costs are manageable, but volume should still be reviewed.

Model-fit matrix

Use this model where capability pays for itself; route away where volume, latency, or budget dominate.

Use caseFitReasonRouting note
Complex reasoning & agentic workflowsExcellentIntelligence score 38 places this model among top performers. Suitable for multi-step analysis and agentic loops.Route complex tasks here; reserve simpler queries for cheaper models.
High-volume chat & customer-facingModerateOutput speed may be a bottleneck for real-time chat at scale.Consider a faster model for latency-sensitive chat.
Latency-sensitive applicationsModerateTTFT 3.00s — latency may be noticeable in interactive use.Consider routing latency-critical paths to faster models.
Cost-sensitive pipelinesModerateOutput pricing at $3.00 is premium. Route high-volume simple tasks to cheaper alternatives.Use only for high-value tasks; route simple queries to budget models.

Cost pressure

Medium

Pricing is moderate — input $0.60, output $3.00. Costs accumulate at volume but are manageable for valuable tasks.

Cheaper substitutes

  • Cheaper routes for predictable extraction, labeling, or summarization

Route away when

  • Predictable high-volume throughput
  • Non-critical classification and extraction

Highlights

Key benchmark results at a glance.

Intelligence

Intelligence position

Open section

Artificial Analysis Intelligence Index · Higher is better · Evaluation results measured independently by Artificial Analysis

Kimi K2.5 (Reasoning) scores 38 (estimated) on the Artificial Analysis Intelligence Index, placing it well above average among other open weight models of similar size (median: 25).

0204060Claude Fable 5 (with fallback): 6060AClaude Fable 5 (with fal…GPT-5.5 (xhigh): 5555AIGPT-5.5 (xhigh)GLM-5.2 (max): 5151?GLM-5.2 (max)Gemini 3.5 Flash: 5050GGemini 3.5 FlashMiniMax-M3: 4444?MiniMax-M3DeepSeek V4 Pro (max): 4444DDeepSeek V4 Pro (max)Kimi K2.6: 4444KKimi K2.6Muse Spark: 4343?Muse SparkKimi K2.5: 3838KKimi K2.5Nemotron 3 Ultra: 3838NNemotron 3 UltraGrok 4.3 (high): 3838xGrok 4.3 (high)gpt-oss-120b (high): 2424AIgpt-oss-120b (high)

Speed

Speed position

Open section

Output tokens per second · Higher is better · Evaluation results measured independently by Artificial Analysis

Kimi K2.5 (Reasoning) generates output at 43.1 tokens per second (based on Kimi's API), which is at the lower end compared to other open weight models of similar size (median: 61.7 t/s).

0113225338450Step 3.7 Flash: 410410?Step 3.7 Flashgpt-oss-120b (high): 314314AIgpt-oss-120b (high)Nemotron 3 Ultra: 250250NNemotron 3 UltraGLM-5.2 (max): 218218?GLM-5.2 (max)Command A+: 196196?Command A+Gemini 3.5 Flash: 192192GGemini 3.5 FlashGrok 4.3 (high): 164164xGrok 4.3 (high)Ring-2.6-1T: 136136?Ring-2.6-1TGLM-4.7: 126126?GLM-4.7DeepSeek V4 Flash (max): 112112DDeepSeek V4 Flash (max)MiniMax-M3: 9999?MiniMax-M3Nex-N2-Pro: 9292?Nex-N2-Pro

Context, Cost & Pricing

Price/value position

Open section

Weighted average cost (USD) per Intelligence Index task · Lower is better · Evaluation results measured independently by Artificial Analysis

Pricing is moderate — input $0.60, output $3.00. Costs accumulate at volume but are manageable for valuable tasks.

$0.00$1.00$2.00$3.00$4.00$5.00DeepSeek V4 Pro (max): $0.04$0.04DDeepSeek V4 Pro (max)gpt-oss-120b (high): $0.06$0.06AIgpt-oss-120b (high)MiniMax-M3: $0.12$0.12?MiniMax-M3Grok 4.3 (high): $0.14$0.14xGrok 4.3 (high)Nemotron 3 Ultra: $0.24$0.24NNemotron 3 UltraKimi K2.6: $0.35$0.35KKimi K2.6GLM-5.2 (max): $0.37$0.37?GLM-5.2 (max)Gemini 3.5 Flash: $0.59$0.59GGemini 3.5 FlashGPT-5.5 (xhigh): $0.86$0.86AIGPT-5.5 (xhigh)Claude Fable 5 (with fallback): $2.75$2.75AClaude Fable 5 (with fal…

Intelligence & Quality

Benchmark scores and quality indices measuring model capability.

AA-Omniscience Index

AA-Omniscience Index (higher is better) measures knowledge reliability and hallucination. It rewards correct answers, penalizes hallucinations, and has no penalty for refusing to answer. Scores range from -100 to 100, where 0 means as many correct as incorrect answers, and negative scores mean more incorrect than correct. · Evaluation results measured independently by Artificial Analysis

Source and methodology are summarized below

Chart source and provenance are listed in Methodology & sources below.

-200204060Claude Fable 5 (with fallback): 4040AClaude Fable 5 (with fal…Gemini 3.5 Flash: 2323GGemini 3.5 FlashGPT-5.5 (xhigh): 2020AIGPT-5.5 (xhigh)Grok 4.3 (high): 1818xGrok 4.3 (high)Kimi K2.6: 6.46.4KKimi K2.6Muse Spark: 4.14.1?Muse SparkGLM-5.2 (max): 44?GLM-5.2 (max)MiMo-V2.5-Pro: 3.63.6?MiMo-V2.5-ProGLM-5.1: 1.91.9?GLM-5.1MiniMax-M3: 1.41.4?MiniMax-M3Kimi K2.5: -8.1-8.1KKimi K2.5

Artificial Analysis Intelligence Index by Open Weights / Proprietary

Artificial Analysis Intelligence Index v4.1 incorporates 9 evaluations: GDPval-AA v2, 𝜏³-Banking, Terminal-Bench v2.1, SciCode, Humanity's Last Exam, GPQA Diamond, CritPt, AA-Omniscience, AA-LCR · Evaluation results measured independently by Artificial Analysis

Source and methodology are summarized below

Chart source and provenance are listed in Methodology & sources below.

0204060Claude Fable 5 (with fallback): 6060AClaude Fable 5 (with fal…GPT-5.5 (xhigh): 5555AIGPT-5.5 (xhigh)GLM-5.2 (max): 5151?GLM-5.2 (max)Gemini 3.5 Flash: 5050GGemini 3.5 FlashMiniMax-M3: 4444?MiniMax-M3DeepSeek V4 Pro (max): 4444DDeepSeek V4 Pro (max)Kimi K2.6: 4444KKimi K2.6Muse Spark: 4343?Muse SparkMiMo-V2.5-Pro: 4242?MiMo-V2.5-ProKimi K2.7 Code: 4242KKimi K2.7 CodeKimi K2.5: 3838KKimi K2.5

Artificial Analysis Intelligence Index

Artificial Analysis Intelligence Index v4.1 incorporates 9 evaluations: GDPval-AA v2, 𝜏³-Banking, Terminal-Bench v2.1, SciCode, Humanity's Last Exam, GPQA Diamond, CritPt, AA-Omniscience, AA-LCR · Evaluation results measured independently by Artificial Analysis

Source and methodology are summarized below

Chart source and provenance are listed in Methodology & sources below.

0204060Claude Fable 5 (with fallback): 6060AClaude Fable 5 (with fal…GPT-5.5 (xhigh): 5555AIGPT-5.5 (xhigh)GLM-5.2 (max): 5151?GLM-5.2 (max)Gemini 3.5 Flash: 5050GGemini 3.5 FlashMiniMax-M3: 4444?MiniMax-M3DeepSeek V4 Pro (max): 4444DDeepSeek V4 Pro (max)Kimi K2.6: 4444KKimi K2.6Muse Spark: 4343?Muse SparkMiMo-V2.5-Pro: 4242?MiMo-V2.5-ProKimi K2.7 Code: 4242KKimi K2.7 CodeKimi K2.5: 3838KKimi K2.5

Artificial Analysis Openness Index: Score

Openness Index assesses model openness on a 0 to 100 normalized scale (higher is more open) · Evaluation results measured independently by Artificial Analysis

Source and methodology are summarized below

Chart source and provenance are listed in Methodology & sources below.

050100Nemotron 3 Ultra: 8383NNemotron 3 UltraMiMo-V2-Flash: 5353?MiMo-V2-FlashDeepSeek V4 Pro (max): 5050DDeepSeek V4 Pro (max)DeepSeek V4 Flash (max): 5050DDeepSeek V4 Flash (max)GLM-5.2 (max): 4444?GLM-5.2 (max)GLM-5.1: 4444?GLM-5.1GLM-4.7: 4444?GLM-4.7GLM-4.6: 4444?GLM-4.6Qwen3 235B A22B 2507: 4444QQwen3 235B A22B 2507gpt-oss-120b (high): 3939AIgpt-oss-120b (high)Kimi K2.5: 3333KKimi K2.5

Intelligence

Artificial Analysis Intelligence Index · Higher is better · Evaluation results measured independently by Artificial Analysis

Source and methodology are summarized below

Chart source and provenance are listed in Methodology & sources below.

0204060Claude Fable 5 (with fallback): 6060AClaude Fable 5 (with fal…GPT-5.5 (xhigh): 5555AIGPT-5.5 (xhigh)GLM-5.2 (max): 5151?GLM-5.2 (max)Gemini 3.5 Flash: 5050GGemini 3.5 FlashMiniMax-M3: 4444?MiniMax-M3DeepSeek V4 Pro (max): 4444DDeepSeek V4 Pro (max)Kimi K2.6: 4444KKimi K2.6Muse Spark: 4343?Muse SparkKimi K2.5: 3838KKimi K2.5Nemotron 3 Ultra: 3838NNemotron 3 Ultra

Speed & Throughput

Output tokens per second and generation throughput.

Output Speed

Output tokens per second · Higher is better · Evaluation results measured independently by Artificial Analysis

Source and methodology are summarized below

Chart source and provenance are listed in Methodology & sources below.

0113225338450Step 3.7 Flash: 410410?Step 3.7 Flashgpt-oss-120b (high): 314314AIgpt-oss-120b (high)Nemotron 3 Ultra: 250250NNemotron 3 UltraGLM-5.2 (max): 218218?GLM-5.2 (max)Command A+: 196196?Command A+Gemini 3.5 Flash: 192192GGemini 3.5 FlashGrok 4.3 (high): 164164xGrok 4.3 (high)Ring-2.6-1T: 136136?Ring-2.6-1TGLM-4.7: 126126?GLM-4.7DeepSeek V4 Flash (max): 112112DDeepSeek V4 Flash (max)

Speed

Output tokens per second · Higher is better · Evaluation results measured independently by Artificial Analysis

Source and methodology are summarized below

Chart source and provenance are listed in Methodology & sources below.

088175263350gpt-oss-120b (high): 314314AIgpt-oss-120b (high)Nemotron 3 Ultra: 250250NNemotron 3 UltraGLM-5.2 (max): 218218?GLM-5.2 (max)Gemini 3.5 Flash: 192192GGemini 3.5 FlashGrok 4.3 (high): 164164xGrok 4.3 (high)MiniMax-M3: 9999?MiniMax-M3GPT-5.5 (xhigh): 8888AIGPT-5.5 (xhigh)Kimi K2.6: 7676KKimi K2.6DeepSeek V4 Pro (max): 7272DDeepSeek V4 Pro (max)Claude Fable 5 (with fallback): 7171AClaude Fable 5 (with fal…Kimi K2.5: 4343KKimi K2.5

Latency & Response Time

Time to first token and end-to-end response latency.

End-to-End Response Time

Seconds to output 500 tokens, including reasoning model 'thinking' time · Lower is better · Evaluation results measured independently by Artificial Analysis

Source and methodology are summarized below

Chart source and provenance are listed in Methodology & sources below.

0s15s30s45s60sGLM-4.6: 10s10s?GLM-4.6MiniMax-M2.7: 9.8s9.8s?MiniMax-M2.7Kimi K2.7 Code: 9.7s9.7sKKimi K2.7 CodeMiMo-V2.5-Pro: 9.5s9.5s?MiMo-V2.5-ProQwen3 235B A22B 2507: 8.1s8.1sQQwen3 235B A22B 2507DeepSeek V4 Pro (max): 7s7sDDeepSeek V4 Pro (max)Kimi K2.6: 6.6s6.6sKKimi K2.6GLM-5.1: 6.3s6.3s?GLM-5.1Nex-N2-Pro: 5.4s5.4s?Nex-N2-ProMiniMax-M3: 5s5s?MiniMax-M3

Latency: Time To First Answer Token

Seconds to first answer token received · Accounts for reasoning model 'thinking' time · Evaluation results measured independently by Artificial Analysis

Source and methodology are summarized below

Chart source and provenance are listed in Methodology & sources below.

0s25s50s75s100sDeepSeek V4 Pro (max): 61s61sDDeepSeek V4 Pro (max)Kimi K2.6: 58s58sKKimi K2.6DeepSeek V4 Flash (max): 50s50sDDeepSeek V4 Flash (max)MiniMax-M2.7: 48s48s?MiniMax-M2.7GLM-5.1: 48s48s?GLM-5.1Kimi K2.7 Code: 43s43sKKimi K2.7 CodeGLM-4.6: 41s41s?GLM-4.6MiMo-V2.5-Pro: 38s38s?MiMo-V2.5-ProQwen3 235B A22B 2507: 33s33sQQwen3 235B A22B 2507Nex-N2-Pro: 22s22s?Nex-N2-Pro

Context, Cost & Pricing

Context window size, cost per task, and token pricing.

Context Window

Context window: tokens limit · Higher is better · Evaluation results measured independently by Artificial Analysis

Source and methodology are summarized below

Chart source and provenance are listed in Methodology & sources below.

0250k500k750k1MGemini 3.5 Flash: 1M1MGGemini 3.5 FlashClaude Fable 5 (with fallback): 1M1MAClaude Fable 5 (with fal…GLM-5.2 (max): 1M1M?GLM-5.2 (max)DeepSeek V4 Pro (max): 1M1MDDeepSeek V4 Pro (max)Grok 4.3 (high): 1M1MxGrok 4.3 (high)MiniMax-M3: 1M1M?MiniMax-M3MiMo-V2.5-Pro: 1M1M?MiMo-V2.5-ProDeepSeek V4 Flash (max): 1M1MDDeepSeek V4 Flash (max)GPT-5.5 (xhigh): 922k922kAIGPT-5.5 (xhigh)Nemotron 3 Ultra: 262k262kNNemotron 3 UltraKimi K2.5: 256k256kKKimi K2.5

Cost per Intelligence Index Task

Weighted average cost (USD) per Artificial Analysis Intelligence Index task, segmented by token type. Lower is better · Evaluation results measured independently by Artificial Analysis

Source and methodology are summarized below

Chart source and provenance are listed in Methodology & sources below.

00.51Claude Fable 5 (with fallback): 0.40.4AClaude Fable 5 (with fal…GPT-5.5 (xhigh): 0.10.1AIGPT-5.5 (xhigh)Gemini 3.5 Flash: 0.10.1GGemini 3.5 FlashGLM-5.1: 00?GLM-5.1GLM-5.2 (max): 00?GLM-5.2 (max)Ring-2.6-1T: 00?Ring-2.6-1TQwen3.5 397B A17B: 00QQwen3.5 397B A17BKimi K2.6: 00KKimi K2.6Nemotron 3 Ultra: 00NNemotron 3 UltraKimi K2.7 Code: 00KKimi K2.7 Code

Cost per Task

Weighted average cost (USD) per Intelligence Index task · Lower is better · Evaluation results measured independently by Artificial Analysis

Source and methodology are summarized below

Chart source and provenance are listed in Methodology & sources below.

$0.00$1.00$2.00$3.00$4.00$5.00Claude Fable 5 (with fallback): $2.75$2.75AClaude Fable 5 (with fal…GPT-5.5 (xhigh): $0.86$0.86AIGPT-5.5 (xhigh)Gemini 3.5 Flash: $0.59$0.59GGemini 3.5 FlashGLM-5.2 (max): $0.37$0.37?GLM-5.2 (max)Kimi K2.6: $0.35$0.35KKimi K2.6Nemotron 3 Ultra: $0.24$0.24NNemotron 3 UltraGrok 4.3 (high): $0.14$0.14xGrok 4.3 (high)MiniMax-M3: $0.12$0.12?MiniMax-M3gpt-oss-120b (high): $0.06$0.06AIgpt-oss-120b (high)DeepSeek V4 Pro (max): $0.04$0.04DDeepSeek V4 Pro (max)

Cost to Run Artificial Analysis Intelligence Index

Cost (USD) to run all evaluations in the Artificial Analysis Intelligence Index · Evaluation results measured independently by Artificial Analysis

Source and methodology are summarized below

Chart source and provenance are listed in Methodology & sources below.

0200400600Claude Fable 5 (with fallback): 508508AClaude Fable 5 (with fal…GPT-5.5 (xhigh): 132132AIGPT-5.5 (xhigh)Gemini 3.5 Flash: 114114GGemini 3.5 FlashGLM-5.1: 4040?GLM-5.1Nemotron 3 Ultra: 3434NNemotron 3 UltraGLM-5.2 (max): 3333?GLM-5.2 (max)Kimi K2.6: 2525KKimi K2.6Qwen3.5 397B A17B: 2525QQwen3.5 397B A17BKimi K2.7 Code: 2424KKimi K2.7 CodeRing-2.6-1T: 2424?Ring-2.6-1T

Pricing: Cache Hit, Input, and Output

Price (USD per M Tokens) · Evaluation results measured independently by Artificial Analysis

Source and methodology are summarized below

Chart source and provenance are listed in Methodology & sources below.

0250k500k750k1MGLM-5.2 (max): 6.16.1?GLM-5.2 (max)GLM-5.1: 6.16.1?GLM-5.1Kimi K2.6: 5.15.1KKimi K2.6Kimi K2.7 Code: 5.15.1KKimi K2.7 CodeQwen3.5 397B A17B: 4.24.2QQwen3.5 397B A17BGrok 4.3 (high): 44xGrok 4.3 (high)Kimi K2.5: 3.73.7KKimi K2.5Nemotron 3 Ultra: 3.63.6NNemotron 3 UltraGLM-4.7: 3.33.3?GLM-4.7Nex-N2-Pro: 3.33.3?Nex-N2-Procache Hitinputoutput

Time per Intelligence Index Task

Weighted average decode time (minutes) per task; excludes TTFT and overhead time · Lower is better · Evaluation results measured independently by Artificial Analysis

Source and methodology are summarized below

Chart source and provenance are listed in Methodology & sources below.

02468MiMo-V2.5-Pro: 6.16.1?MiMo-V2.5-ProGLM-4.6: 5.95.9?GLM-4.6MiniMax-M2.7: 5.55.5?MiniMax-M2.7Qwen3.5 397B A17B: 5.15.1QQwen3.5 397B A17BKimi K2.7 Code: 55KKimi K2.7 CodeGLM-5.1: 4.94.9?GLM-5.1Nex-N2-Pro: 4.84.8?Nex-N2-ProClaude Fable 5 (with fallback): 4.84.8AClaude Fable 5 (with fal…Qwen3 235B A22B 2507: 4.44.4QQwen3 235B A22B 2507MiniMax-M3: 3.93.9?MiniMax-M3

Architecture & Scale

Model size, parameters, and architectural details.

Model Size: Total and Active Parameters

Comparison between total model parameters and parameters active during inference · Evaluation results measured independently by Artificial Analysis

Source and methodology are summarized below

Chart source and provenance are listed in Methodology & sources below.

0250k500k750k1MDeepSeek V4 Pro (max): 1.6k1.6kDDeepSeek V4 Pro (max)MiMo-V2.5-Pro: 1k1k?MiMo-V2.5-ProKimi K2.5: 1k1kKKimi K2.5Kimi K2.6: 1k1kKKimi K2.6Kimi K2.7 Code: 1k1kKKimi K2.7 CodeRing-2.6-1T: 1k1k?Ring-2.6-1TGLM-5.2 (max): 753753?GLM-5.2 (max)GLM-5.1: 744744?GLM-5.1DeepSeek R1 (Jan): 685685DDeepSeek R1 (Jan)Nemotron 3 Ultra: 550550NNemotron 3 Ultrapassive Paramsactive Params

Specifications

Technical Specifications

Kimi K2.5 (Reasoning) scores 38 (estimated) on the Artificial Analysis Intelligence Index, placing it well above average among other open weight models of similar size (median: 25). Kimi K2.5 (Reasoning) generates output at 43.1 tokens per second (based on Kimi's API), which is at the lower end compared to other open weight models of similar size (median: 61.7 t/s). Kimi K2.5 (Reasoning) costs $0.60 per 1M input tokens (somewhat higher than average, median: $0.59) and $3.00 per 1M output tokens (somewhat higher than average, median: $2.20), based on Kimi's API.

Kimi K2.5

Model type
Open weights
Reasoning
Yes
Input modalities
Kimi K2.5 (Reasoning) supports text, image, and video input.
Output modalities
Kimi K2.5 (Reasoning) supports text output.
Context window
260k tokens
Open weights / source
Yes, Kimi K2.5 (Reasoning) is open weights. The model weights are publicly available and can be downloaded for self-hosting.
Parameters
Kimi K2.5 (Reasoning) has 1 trillion parameters (32 billion active).
Active parameters
Kimi K2.5 (Reasoning) is a Mixture of Experts (MoE) model with 1 trillion total parameters, but only 32 billion active parameters are used during inference.
License
Kimi K2.5 (Reasoning) is released under the Modified MIT License license. This license allows commercial use.
API availability
Yes, Kimi K2.5 (Reasoning) is available via API through 13 providers.

Methodology & Provenance

This page is rendered from the normalized profile and page JSON for Kimi K2.5.

Benchmark values are preserved as normalized; only layout, disclosure ordering, and typography are adjusted for readability.

Frequently Asked Questions