K

Kimi K3

Moonshot AI · Released Jul 2026

Conditional

Strong model now under active White House investigation for industrial-scale IP theft. US Treasury threatens sanctions — legal and operational risk is material. Technical benchmarks unchanged.

Is it right for you?

Good for

  • Front-end and 3D work with image feedback — Arena WebDev #1 (1679, preliminary)
  • Long-horizon autonomous coding — MiniTriton compiler, 48-hour chip design demonstrations
  • Multi-agent orchestration — K3 Swarm Max variant for large-scale parallel processing
  • Large-context knowledge work — 1M window for whole-repo and multi-document research

Not good for

  • Precise-execution tasks — still trails Fable 5 and GPT 5.6 Sol
  • Self-hosted deployment — requires 64+ accelerators, weights not yet downloadable
  • Cost-sensitive workloads — AA measured high output verbosity driving up task costs

Pricing

Input

$3.00 / 1M

Output

$15.00 / 1M

Context

1M ctx, $0.30 cached. Weights live (Jul 27).

View full pricing

Benchmarks

BenchmarkScoreSource
DeepSWE67.5 Source
Terminal-Bench 2.188.3 Source
FrontierSWE81.2 Source
Program Bench77.8 Source
SWE Marathon42.0 Source
Arena WebDev1679 (#1, preliminary) Source
AA Intelligence Index57 Source

Verdict history

Jul 23, 2026
conditional -> conditional (rationale updated) — On July 22, 2026, White House OSTP Director Michael Kratsios publicly accused Moonshot AI of industrial-scale distillation of Anthropic's Fable 5 to train Kimi K3. Treasury Secretary Bessent threatened sanctions and Entity List designations.

Verification log

  • Pricing— No changes

    Automated agent

  • Pricing— Updated

    Automated agent

    Kimi K3 open weights went live on Hugging Face Jul 26 evening EDT (one day ahead of Jul 27 schedule). Model is now truly open-weight under Modified MIT license. Pricing unchanged at $3/$15. This is a material status change from "weights not yet downloadable" to "weights live on Hugging Face".

  • Pricing— No changes

    Automated agent

How we evaluate