Moonshot AIChina· released Jul 2026· open weights

Kimi K3

Name: Kimi K3
Author: Moonshot AI

Ranked #8, 20.4% behind the frontier on average.

Moonshot AI, China. One of 5 models from this lab on the Index.

Third-party publishedMedium confidence (7+ comparable tests, 3+ Index categories)

Available in Theo as Theo Open Max

Theo orchestrates it for the steps it does best and always shows which engine answered.

Use it in Theo

Snapshot September 8, 2026

OpenCharts Index

79.6

Ranked #8 (6 to 10) of 36 · 20.4% behind the frontier on average · 18 comparable tests · 3 counting categories

Drop any single test and the Index lands between 77.2 and 83.5, anywhere from #6 to #10. Neighbours inside that range are ties.

Needs more results (3 of 8 so far)

Averaging 60.2 on the 3 it has; the score opens at 4 of 8.

Distance to today's frontier on the hardest tests. Not a measure of intelligence.

What it's known for

Kimi K3, in 18 results.

Strongest in coding (83.0) and reasoning (80.5). It trails the frontier most in math, 24.7% behind on average. No comparable results yet in multimodal and long context. Between 20% and 25% behind the best published results, on average. Drop any single test and it would sit anywhere from #6 to #10.

Strongest category

Coding

Writing, fixing and shipping real code, judged by tests or users.

83.0

SciCode58.7% · 95% of best
WeirdML82.6% · 89% of best
CursorBench60.8% · 83% of best
FrontierCode44.2% · 83% of best
Code Arena (WebDev)1,674 · 66% win rate vs leader

Mean of 5 of the category's 7 comparable tests. Counts toward the Index.

Best single result

Mock AIME 2024–2025

Solve competition mathematics problems that have a single exact integer answer.

97.2%

97.2% of the best (100.0%) · #15 of 78.

Olympiad-style problems need long, careful chains of reasoning, and there is no partial credit.

Where it trails

FrontierMath Tier 4

Solve the hardest FrontierMath tier: problems that take expert mathematicians days.

39.0%

60.0% behind the best published result (97.6%) · #16 of 51.

The deepest end of the set. Progress here signals genuinely new capability.

Not measured yet

No comparable result yet in these 2 categories. A blank is a blank, never a zero, and it does not lower the Index.

Multimodal
Long context

Covered, but not averaged into the Index:

Knowledgesingle-test category, shown beside the Index · 66.9
Agentic2 of 8 tests · needs 4 · mean 79.0
Human preferencesingle-test category, shown beside the Index · 94.7

Every score, with its source

See it at work

What Kimi K3 was asked to do, and how it did.

Every test it has taken, by category, with the kind of task it faced, who produced the number and where the result landed against the best published one. The examples are original and illustrative, never items from the datasets themselves.

Human preference · mean 94.7 across 1 comparable test · a single-test category, shown beside the Index and never averaged into it

Text ArenaLeaderboard

Answer the same prompt as a rival model; real users vote blind on which answer they prefer.

1,489

47% expected win rate against the board leader (1,507)
#10 of 92

max · board 2026-09-02

Source: Arena

Every result is read against the best published one on its test: percent scores as a share of the best above chance, Arena ratings as the expected win rate against the board leader, open-ended values on a log scale. Sitting a harder exam never lowers a model. Every number links to the publisher that produced it.

Head to head

Kimi K3 against whoever you pick.

Choose a rival. Every counted test both have taken appears side by side as a share of the best published result on that test, with the category means above. A win is a higher share; a gap under 2 points is a tie, because one item on a 45-problem exam is 2.2 points and publishers' own standard errors are of that order.

Kimi K3

#8·Index 79.6B

1 : 13

wins · 0 ties · 14 shared

GPT-6 Astra

#1·Index 99.0A+

GPT-6 Astra comes out ahead on 13 of the 14 tests both have taken.

Category means

Reasoning
80.5
99.6
Knowledge
66.9
100.0
Coding
83.0
97.7
Math
75.3
99.8
Agentic
79.0
98.5
Human preference
94.7
—

Shared tests · biggest gaps first

FrontierMath Tier 4Math
39.0%
97.6%
−60.0
ARC-AGI-2Reasoning
60.4%
95.0%
−36.4
Code Arena (WebDev)Coding
1,674
1,797
−33.9
SimpleQA VerifiedKnowledge
50.6%
75.6%
−33.1
CritPtReasoning
23.4%
31.7%
−25.8
FrontierMath (Tiers 1–3)Math
72.2%
93.7%
−23.0
FrontierCodeCoding
44.2%
53.3%
−17.0
APEX-AgentsAgentic
39.3%
46.7%
−15.6
ProofBenchMath
87.0%
99.0%
−12.0
WeirdMLCoding
82.6%
92.9%
−11.1
ARC-AGI-1Reasoning
94.5%
98.5%
−4.1
GPQA DiamondReasoning
93.1%
95.8%
−3.7
SciCodeCoding
58.7%
56.5%
+3.6
Mock AIME 2024–2025Math
97.2%
100.0%
−2.8

Bars are distances to the best published result on each test (percent scores above chance, Arena ratings as win rate against the leader, open-ended values on a log scale). Raw values as the publishers report them.

Open GPT-6 Astra

Scores

Every test, every source

The best published run per benchmark, the raw value as the publisher reports it, who produced the number, how close it comes to the best published result on that test, and where the model sits among every model scored on it. Percent scores are read above chance, Arena ratings as a win rate against the board leader, open-ended values on a log scale. A test with too few models to compare against is shown but not counted.

Category means

Reasoning80.5 · 5/6
Knowledgebeside66.9 · 1/1
Coding83.0 · 5/7
Math75.3 · 4/5
Agentic79.0 · 2/8
Multimodal—
Human preferencebeside94.7 · 1/1
Long context—

Mean · tests taken of the category's comparable tests. A faded category either rests on fewer than half of them or is a single-test category (marked “beside”), which is shown next to the Index and never averaged into it.

GPQA DiamondEpoch-runReasoning

93.1%

96.3% of the best (95.8%), above chance
86th percentile of 78 · n = 198

max

Kimi K3

Kimi K3, in 18 results.

What Kimi K3 was asked to do, and how it did.

Kimi K3 against whoever you pick.

Every test, every source

Compare with its neighbours.

The best models, ranked here, working inside Theo.