MoonshotOctober 23, 2026Official source

Kimi K3.1 Pro

Name: Kimi K3.1 Pro
Price: 2 USD
Author: Moonshot

Post-trained version of Kimi K3 with a one-million-token context window. API only for now: Moonshot plans to publish the weights on November 4.

Singularity Index

64.8

RANK 2 / 42

SI v3 · 100% coverage

Share of best score

Reasoning

30%

HLE

Coding

25%

SciCode

Agentic

25%

Terminal-Bench 4.0

Multimodal

10%

MMMU-Pro

Long Context

10%

AA-LCR

SI v3 uses weighted percentages from one Artificial Analysis configuration. Ranking requires all five benchmarks. Partial scores average only available weights. Scores are not comparable to SI v2 or a probability of reaching singularity.

STRENGTHS AND TRADEOFFS

Compared with 57 models released within six months of it, across 3 benchmarks where scores spread out.

Strongest on

Terminal-Bench 4.0 #1 of 5367.2%
HLE #10 of 5252.5%
AA Intelligence v5.0 #10 of 2946

Cost and access

$4 per million tokensCostlier than 71% of 28 priced contemporaries (3:1 input to output)
1M-token context

BENCHMARKS

Benchmark	Score	Rank
HLEArtificial Analysis Challenging multidisciplinary questions evaluated by Artificial Analysis	52.5%	#10 / 102

ADDITIONAL EVALUATIONS

Independent HLE, SciCode, Terminal-Bench 4.0, MMMU-Pro, and AA-LCR contribute to SI v3. Provider reports and other evaluations are display-only.

Benchmark / source	Score
HLE Artificial Analysis Kimi K3.1 Pro	52.5%
MMMU-Pro Artificial Analysis Kimi K3.1 Pro	83.9%
Terminal-Bench 4.0 Artificial Analysis Kimi K3.1 Pro	67.2%
SciCode Artificial Analysis Kimi K3.1 Pro	60.8%
AA-LCR Artificial Analysis Kimi K3.1 Pro	86.3%
CritPt Artificial Analysis Kimi K3.1 Pro	24.6%