vibecodingnav
中/EN
Models/kimi-k2-7-code

Kimi K2.7 Code

14th on the current Next.js board, 74% without AGENTS.md and 90% with it. 59th on ReactBench v1.2 at 23.6%.

Moonshot · Proprietary · Fetched 2026-10-10

LMArena WebDev
1473
#65 / 142 · ±9
LMArena Image-to-WebDev
1513
#40 / 64 · ±20
Next.js Agent Evals
74.0%
#14 / 14

Benchmark scores

Boards older than 90 days are not counted as recommendations
BenchmarkSettingScoreRankCostAs of
LMArena WebDev—1473±9#65—2026-10-08
LMArena Image-to-WebDev—1513±20#40—2026-10-09
Next.js Agent EvalsOpenCode74.0#14$0.1392026-09-25
ReactBench—23.620.3–26.9#59$1.272026-10-10
DeepSWE v1.1mini-swe-agent30.530–31#26$2.822026-09-22
SWE-bench Verified—This board has stopped updating——2026-02-26
Aider Polyglot—This board has stopped updating——2025-11-20

Use it or not

Why
  • The Next.js docs move pass@4 from 74% to 90%.
When not
  • The baseline is last on the current board. Do not merge the two evals into one score.

Appears in these scenarios

Collected from scenario boards
Build a Web frontend#14Next.js 74.0%Build an Android app#26DeepSWE 30.5%Write CLI, scripts, and DevOps#26DeepSWE 30.5%

Peers

ModelWebDevPrice $/M
Kimi K2.7 CodeCurrent1473—MiniMax M31482—