Kimi Code Bench v2

Kimi K3.1 Pro 77.1% leads since October 2026

In use

Coding agents complete Moonshot's realistic tasks from internal engineering work, production incidents, and open-source projects in over ten languages.

In-house: Moonshot never released the tasks, and only Moonshot reports scores on it.

Made by: Moonshot AI
Released: June 2026
Kind: Fixing real repositories
Level: Expert
Scored: Checked automatically
Answers: Code
Tasks: Kept private

Strongest model tested: Kimi K3.1 Pro #16 on the capability index
Last in a launch post: October 2026
Last model added: October 2026

Models scored in lab cards

Launch posts list it, from 1 lab

System card reports it, from 1 lab

Kimi K3.1 Pro leads with 77.1%

Best score per model as labs report them in their own cards. Labs pick their own settings, so these compare less cleanly than independent runs.

Bars Table

1 Kimi K3.1 Pro 77.1%
2 Kimi K3 72.9%
3 Kimi K2.7 Code 62%

Source: Labs’ model cards

#	Model	Score	Date	Source
1	Kimi K3.1 Pro	77.1%	Oct 2026	Lab model card
2	Kimi K3	72.9%	Jul 2026	Lab model card
3	Kimi K2.7 Code	62%	Jun 2026	Lab model card

3 models have held number one in labs’ own numbers since its release in June 2026

Lab claims count from the day the card came out, independent runs from the model’s release.

Moonshot AI

Lab claims 3 leaders

Kimi K3.1 Pro 77.1%

Source: Labs’ model cards

Scores over time

All the scores we have, one dot each, so a model tested at several settings appears more than once.

Reported by the lab in its model card Best so far

Source: Labs’ model cards

One lab put it in a launch post in the last six months

Launch posts from every lab that has listed it, one tick per post. Tall colored ticks include it, and short gray ones leave it out.

Moonshot AI

Date	Launch post	Score	As the post writes it
Oct 24, 2026	Kimi K3.1 Pro	77.1	Kimi Code Bench 2.0 (Internal)
Jul 16, 2026	Kimi K3	72.9	Kimi Code Bench 2.0
Jun 12, 2026	Kimi K2.7 Code	62	Kimi Code Bench v2

One system card reports it

Besides launch posts, we read the system cards, model cards and technical reports labs publish with a release. They report far more, including safety and frontier-risk tests.

Date	Document	Score	As the document writes it	Section
Jul 27, 2026	Kimi K3 technical report	72.9	Kimi Code Bench 2.0 with Kimi Code harness, as in Figure 1. 73.7 with Claude Code harness (Table 3)	Capabilities

Kimi K3.1 Pro leads with 77.1%

3 models have held number one in labs’ own numbers since its release in June 2026

Scores over time

One lab put it in a launch post in the last six months

One system card reports it

More in Coding