Inference
EN
Moonshot AI

Kimi K2.5 (Reasoning)

Kimi K2.5 (Reasoning) is a reasoning model from Moonshot AI in the Kimi K2.5 family. 21 benchmarks count toward its index, in 7 categories.

  • Open weights
  • Reasoning
  • Input: text
  • Released 2026-01
RankedEnough results for an index.
IndexAll categories, out of 100.53.0±4.0Median 51.0
CoverageIndex weight that has results.95%
SpeedOutput tokens per second.—Median 67/s
First tokenSeconds to the first output token.—Median 1.3 s
Input / 1MUS dollars per 1M input tokens.$0.60Median $0.60
Output / 1MUS dollars per 1M output tokens.$3.00Median $2.55
ContextMaximum tokens in one request.128KMedian 262K
EloLMArena rating and rank.—Median 1422

Run locally

1026.88B parameters
16-bit
2157.4 GB
8-bit
1146.6 GB
4-bit
654.7 GB
Weights
Hugging Face (opens in a new tab) ↗

Memory need with a short context. The 4-bit copy needs more than 128 GB. Compare local models

Capabilities

Index per category, out of 100.
Out of 100
AgenticMulti-step tasks with tools.7 counted50.6123 / 218
CodingCode writing and repair.2 counted56.090 / 209
ReasoningLogic problems and puzzles.3 counted52.1101 / 227
MultimodalTasks with images and text.2 counted53.958 / 84
KnowledgeFacts and expert knowledge.4 counted55.593 / 268
MultilingualTasks in many languages.None counted——
InstructionTasks with strict prompt rules.1 counted54.6—
MathMath problems.2 counted52.277 / 168

Results

21 counted
BenchmarkThe benchmark name.ResultThe score from the source.PlacePlace among results on this benchmark.
τ²-Bench Tool-Agent-User EvaluationAgentic95.9%13 / 149
Massive Multitask Language Understanding ProfessionalKnowledge87.1%7 / 75
American Invitational Mathematics Examination 2025Math96.1%6 / 53
Artificial Analysis GPQA DiamondKnowledge87.9%53 / 199
Harvard-MIT Mathematics Tournament February 2025Math95.4%4 / 15
Graduate-Level Google-Proof Q&AKnowledge87.6%41 / 141
Artificial Analysis IFBenchInstruction70.2%44 / 146
LiveCodeBench v6Coding85.0%19 / 59
Artificial Analysis Long Context ReasoningReasoning78.0%70 / 217
Software Engineering Benchmark VerifiedCoding76.8%38 / 114
Artificial Analysis Humanity's Last ExamKnowledge30.7%74 / 222
Artificial Analysis Omniscience AccuracyKnowledge35.2%65 / 164
Artificial Analysis MMMU-ProMultimodal75.4%49 / 113
Massive Multi-discipline Multimodal Understanding ProMultimodal78.5%23 / 53
LongBench v2Reasoning61.0%11 / 25
Critical Physics TasksReasoning3.1%95 / 214
GDPval-AA normalizedAgentic17.0%83 / 133
Terminal-Bench 2.0 · BenchLMAgentic50.8%28 / 41
OSWorld-VerifiedAgentic63.3%29 / 37
CharXiv ReasoningMultimodal77.5%40 / 49
APEX-Agents-AAAgentic11.5%25 / 30
BrowseCompAgentic60.6%53 / 63
Gert Labs Composite Game BenchmarkAgentic32.6%46 / 52

Sources

BenchLMCC BY-NC 4.0

History

5 releases
Index53.9 → 53.0
23 Oct 202612 Nov 2026
Output / 1M$3.00 → $3.00
23 Oct 202612 Nov 2026

Questions

How much does Kimi K2.5 (Reasoning) cost?
$0.60 per 1M input tokens, $3.00 per 1M output tokens.
How good is Kimi K2.5 (Reasoning)?
Its index is 53.0 out of 100. That is place 117 of 283 ranked models.
What is Kimi K2.5 (Reasoning) best at?
Knowledge: 55.5 out of 100, place 93 of 268.
What is the context window of Kimi K2.5 (Reasoning)?
128K tokens.
Is Kimi K2.5 (Reasoning) open source?
Yes. Its weights are published.
When was Kimi K2.5 (Reasoning) released?
January 2026.

Badge

Kimi K2.5 (Reasoning) on Inference360: place 117 of 283, index 53.0
HTML
<a href="https://inference360.ai/models/kimi-k2-5-reasoning"><img src="https://inference360.ai/badges/kimi-k2-5-reasoning.svg" alt="Kimi K2.5 (Reasoning) on Inference360: place 117 of 283, index 53.0" height="20"></a>
Markdown
[![Kimi K2.5 (Reasoning) on Inference360: place 117 of 283, index 53.0](https://inference360.ai/badges/kimi-k2-5-reasoning.svg)](https://inference360.ai/models/kimi-k2-5-reasoning)

Compare with

GPT-5 (medium)53.0Muse Glimmer 30B53.0GPT-5.1-Codex52.9Grok 4.2052.8Kimi K2.552.2

Lower price

dots3-note Preview63.4 · FreeApodex 1.1 Mini55.7 · FreeNemotron 3 Ultra50.8 · Free

More from Moonshot AI

All models
Kimi K2.552.2Kimi K2 Thinking48.2Kimi K2.658.2Kimi LatestUnrankedKimi K2.7 Code57.7