KimiReleased Jul 11, 2025
Open weights
kimi-k2
Its best results are in Mathematical Reasoning (#107 of 176) and Programmatic Reasoning (#122 of 177); weakest in Reliability (#152 of 177). Three of seven dimensions are estimated, because no results are published yet.
Composite IQ ?
91
Effective cost
$2.38/1M
$0.57 in · $2.30 out
Speed
46tok/s
13s typical response
Context
Unverified
Not in publisher docs
Evidence
4of 7 dims
14 of 39 benchmarks
IQ profile
Seven dimensions, averaged equally into Composite IQ
MeasuredEstimated95% range
#1 in dimensionTop-10 median
IQ dimensions
Ranks are among models scored in each dimension. Ranges are model-based estimates, not calibrated confidence intervals.
DimensionIQPosition · 70–160RankEvidencevs Top 10
kimi-k295% range (model-based, uncalibrated)Top-10 median#1 in dimension100 = average human
Benchmark results
14 of 39 Composite IQ benchmarks have a published result for kimi-k2. Benchmarks without a result stay listed so gaps are visible.
Abstract Reasoning IQ 72 · estimatedNo published results0/5
No published results yet for Abstract Reasoning. The 72 shown is estimated from kimi-k2's other dimensions and will be replaced as results come in. How estimates work
ARC-AGI-3ARC-AGI-2ARC-AGI-1Chess PuzzlesMystery Game Puzzles
Mathematical Reasoning IQ 105 · measuredBest: #17 of 64 on MathArena4/5
BenchmarkScorePercentileRankLeader · source
FrontierMath Tier 1-3
Epoch AI FrontierMath Tiers 1-3 v2 · Sep 22
21.4%
#71 / 84
Leader gpt-6-astra · 93.7%
Epoch AI FrontierMath Tiers 1-3 v2 · Sep 22
Epoch AI FrontierMath Tiers 1-3 v2 · Sep 22
FrontierMath Tier 4
Epoch AI FrontierMath Tier 4 v2 · Sep 22
0.0%
#72 / 72
Leader gpt-6.1-sol · 100%
Epoch AI FrontierMath Tier 4 v2 · Sep 22
Epoch AI FrontierMath Tier 4 v2 · Sep 22
Not yet reported
ProofBench 1.1
Academic Reasoning IQ 103 · measuredBest: #113 of 147 on GPQA Diamond4/6
BenchmarkScorePercentileRankLeader · source
GPQA Diamond
Artificial Analysis v4.3.2 · Oct 27
76.6%
#113 / 147
Leader gpt-6-astra · 96.1%
Artificial Analysis v4.3.2 · Oct 27
Artificial Analysis v4.3.2 · Oct 27
SciCode
Artificial Analysis model leaderboard · Sep 10
34.5%
#146 / 167
Leader opus-5.5 · 66.9%
Artificial Analysis model leaderboard · Sep 10
Artificial Analysis model leaderboard · Sep 10
Humanity's Last Exam
Artificial Analysis v4.3.2 · Oct 27
7.4%
#151 / 167
Leader opus-5.5 · 61.4%
Artificial Analysis v4.3.2 · Oct 27
Artificial Analysis v4.3.2 · Oct 27
CritPt
Artificial Analysis v4.3.2 · Oct 27
0.0%
#148 / 159
Leader Muse Wonder Full · 32.9%
Artificial Analysis v4.3.2 · Oct 27
Artificial Analysis v4.3.2 · Oct 27
Not yet reported
MMLU-ProMMMU-Pro
Programmatic Reasoning IQ 100 · measuredBest: #54 of 61 on LiveCodeBench1/6
BenchmarkScorePercentileRankLeader · source
Not yet reported
IOI V2Terminal-Bench 4.0ProgramBenchSWE-rebenchMirrorCode
Computer Use IQ 70 · estimatedNo published results0/5
No published results yet for Computer Use. The 70 shown is estimated from kimi-k2's other dimensions and will be replaced as results come in. How estimates work
BrowseCompOSWorld-VerifiedMCP AtlasArena.ai Agent ArenaAgents' Last Exam
Reliability IQ 99 · measuredBest: #121 of 158 on AA Omniscience5/6
BenchmarkScorePercentileRankLeader · source
AA Omniscience
Artificial Analysis v4.3.2 · Oct 27
-28.3
#121 / 158
Leader opus-5.5 · 46.4
Artificial Analysis v4.3.2 · Oct 27
Artificial Analysis v4.3.2 · Oct 27
SimpleQA Verified
Epoch AI SimpleQA Verified · Sep 22
31.6%
#64 / 80
Leader Gemini 4 Argon · 76.0%
Epoch AI SimpleQA Verified · Sep 22
Epoch AI SimpleQA Verified · Sep 22
AA Long Context Reasoning v1.1
Artificial Analysis v4.3.2 · Oct 27
53.0%
#134 / 154
Leader kimi-k3 · 88.7%
Artificial Analysis v4.3.2 · Oct 27
Artificial Analysis v4.3.2 · Oct 27
BullshitBench v2
Full source leaderboard reconciliation · Oct 23
10.0%
#75 / 83
Leader opus-4.8 · 94.0%
Full source leaderboard reconciliation · Oct 23
Full source leaderboard reconciliation · Oct 23
IFBench
Artificial Analysis v4.3.2 · Oct 27
41.5%
#88 / 96
Leader minimax-m3 · 82.9%
Artificial Analysis v4.3.2 · Oct 27
Artificial Analysis v4.3.2 · Oct 27
Not yet reported
FACTS Grounding
Knowledge Work IQ 90 · estimatedNo published results0/6
No published results yet for Knowledge Work. The 90 shown is estimated from kimi-k2's other dimensions and will be replaced as results come in. How estimates work
AA-Briefcase v1.1GDPval-AA v2.1AutomationBench-AAGDP.pdfAA-AnalystAgentEnterpriseOps-Gym-AA
Release & availability
Release details on record for this model.
- Released
- Developer
- Kimi
- Weights
- Open weightsDownloadable and self-hostable
- API price
- $0.57 in · $2.30 outper 1M tokens · full pricing
About this score
How much of kimi-k2's IQ is measured, and how the rest is estimated.
14/ 39benchmark results
4/ 7dimensions measured
3estimated
0/5Abstract
4/5Math
4/6Academic
1/6Coding
0/5Computer Use
5/6Reliability
0/6Knowledge Work
Bar height = share of benchmarks with resultsEstimatedNo score
Measured results are shown as published
Every score in the benchmark tables comes from its source unchanged. Composite IQ averages all seven dimensions equally.
Estimates fill the gaps
Abstract Reasoning, Computer Use and Knowledge Work have no published results yet, so their IQ is estimated from kimi-k2's other dimensions.
A missing result is not a zero
Benchmark versions and evaluation harnesses can differ between sources, so small gaps between models may not be meaningful.
Compare kimi-k2
Side-by-side pages for the models it is most often weighed against.
#1 overall
gpt-6.1-sol
IQ141 50 higherEffective cost$6.30 2.6× pricier
Kimi #1
Kimi K3.1 Pro
IQ132 41 higher
Closest IQ
mistral-large-3
IQ91 same IQEffective cost$0.44 5.4× cheaper
Most similar profile
deepseek-r1
IQ93 2 higherEffective cost$3.12 1.3× pricier
Similar price
DeepSeek V4.1 Flash 1017
IQ128 37 higherEffective cost$2.42 about the same
#2 overall
opus-5.5
IQ141 50 higherEffective cost$35.73 15.0× pricier
kimi-k2Comparison model
Frequently asked questions
Quick answers about kimi-k2, drawn from the same data as the rest of this page.
How AI IQ scores models →Get this data from the API →- What is the IQ of kimi-k2?
- kimi-k2 has an estimated IQ of 91, which ranks #159 out of 176 public models on the AI IQ leaderboard. The score combines seven equally weighted reasoning dimensions using measured benchmark results and conservative estimates for missing coverage.
- What is kimi-k2 best at?
- Among the dimensions with published benchmark results, kimi-k2 ranks highest in Mathematical Reasoning (IQ 105, #107 of 176) and Programmatic Reasoning (IQ 100, #122 of 177). Its lowest-ranked measured dimension is Reliability (IQ 99, #152 of 177).
- What is the EQ of kimi-k2?
- kimi-k2 doesn't have an EQ score yet. EQ (emotional reasoning) is calculated from EQ-Bench 4 and AttuneBench, and neither has published a result for kimi-k2. AI IQ doesn't estimate EQ from other benchmarks, so the score appears once one of those results is published.
- Which benchmarks has kimi-k2 been tested on?
- kimi-k2 has published results on 14 of the 39 benchmarks behind AI IQ's Composite IQ, including MathArena, AIME, AA Omniscience, GPQA Diamond and SimpleQA Verified.
- How much does kimi-k2 cost?
- kimi-k2 costs $0.57 per million input tokens and $2.30 per million output tokens. Adjusted for how many tokens it actually consumes on real tasks, its effective cost works out to $2.38 per million input-plus-output tokens.
- Is kimi-k2 open source?
- Yes — kimi-k2 is an open-weights model: its trained weights are published, so it can be downloaded and self-hosted.
- Who developed kimi-k2?
- kimi-k2 was developed by Kimi, based in China.
- When was kimi-k2 released?
- kimi-k2 was publicly released on July 11, 2025.
- How large is the context window of kimi-k2?
- AI IQ hasn't verified an input context limit for kimi-k2 from its developer's documentation, so none is listed rather than guessing.
- How fast is kimi-k2?
- In independent measurements, kimi-k2 generates about 46 tokens per second, with a typical response time of about 13 seconds.