Kimi K2.5 (Reasoning)

Kimi K2.5 (Reasoning) is a reasoning model from Moonshot AI in the Kimi K2.5 family. 19 benchmarks count toward its score, in 7 categories.

availableShows if the model has enough results for an index.
IndexOverall score out of 100.54.6 ±4.2
CoverageShare of the index weight with results.95%
SpeedOutput tokens per second.
Input / 1MUS dollars per 1M input tokens.$0.6
Output / 1MUS dollars per 1M output tokens.$3
ContextMaximum tokens in one request.128K
EloLMArena rating and rank.N/A

The index is a score out of 100. The ± range shows how much it can change.

CapabilitiesScore per category, out of 100.

Out of 100
AgenticMulti-step tasks with tools.
51.6
CodingCode writing and repair.
55.5
ReasoningLogic problems and puzzles.
53.5
MultimodalTasks with images and text.
57.2
KnowledgeFacts and expert knowledge.
57.1
MultilingualTasks in many languages.
N/A
InstructionTasks with strict rules in the prompt.
60.3
MathMath problems.
52.2

Results

19 counted
BenchmarkThe test name.CategoryThe capability that the test measures.ResultThe score from the publisher.IndexThis result as a score out of 100.RunThe settings of the run.DateDate of the result.Published byThe source of the result.
American Invitational Mathematics Examination 2025Math96.1%52.2Mathematical Association of America
τ²-Bench Tool-Agent-User EvaluationAgentic95.9%67.6Victor Barres et al.
Artificial Analysis GPQA DiamondKnowledge87.9%58.6Artificial Analysis
Graduate-Level Google-Proof Q&AKnowledge87.6%59.1David Rein et al.
Massive Multitask Language Understanding ProfessionalKnowledge87.1%57.8Yubo Wang et al.
Massive Multi-discipline Multimodal Understanding ProMultimodal78.5%55.1MMMU-Pro authors
Artificial Analysis Long Context ReasoningReasoning78.0%62.2Artificial Analysis
Software Engineering Benchmark VerifiedCoding76.8%59.1Carlos E. Jimenez et al.
Artificial Analysis MMMU-ProMultimodal75.4%59.2Artificial Analysis
Artificial Analysis IFBenchInstruction70.2%60.3Artificial Analysis
BrowseCompAgentic60.6%50.8OpenAI
Artificial Analysis Coding IndexCoding46.8%51.9Artificial Analysis
Artificial Analysis Omniscience AccuracyKnowledge35.2%57.4Artificial Analysis
Gert Labs Composite Game BenchmarkAgentic32.6%41.5Gert Labs
Artificial Analysis Humanity's Last ExamKnowledge30.7%57.8Artificial Analysis
Artificial Analysis Intelligence IndexKnowledge23.5%51.7Artificial Analysis
GDPval-AA normalizedAgentic16.2%48.0Artificial Analysis
APEX-Agents-AAAgentic11.5%50.2Artificial Analysis / Mercor
Critical Physics TasksReasoning3.1%44.9Artificial Analysis

19 benchmarks count, from 19 of 19 results. A grey row does not count. Too few models took that benchmark.

Sources

BenchLM benchmark aggregationUsed with attribution; per-benchmark results credited to their original authors

Same level, lower price

Qwen3.8-Flash-Next66.4 · Freedots3-note Preview65.5 · FreeApodex 1.1 Mini59.1 · Free

More from Moonshot AI

Kimi K372.0Kimi K2.660.9Kimi K2.7 Code59.5Kimi K2.553.3Kimi K2.5 Thinking53.8