Use · Kimi K2
Is it good enough?
Ownership levelPartialnone·limited·partial·substantial·fullAnalytical input C ยท 62.8/100
This page is a projection of the one entry record, the Reliability factor that Use covers. The full verdict is set by all four factors together, floor-weighted so the weakest caps the whole.
Which domain expands which factor
- AssessUse & modify + Transparency
- ImplementData control + Reliability
- UseReliability
- SupportTransparency
Capabilities & modalities
Kimi K2 is a text-focused, frontier-adjacent general model with strong agentic tool use, coding
and reasoning. Per Moonshot's technical report and independent evaluation (ev-perf,
ev-tech-report) it is among the strongest open-weight models on agentic and coding benchmarks.
It is designed as an "open agentic intelligence". Choose it for frontier-adjacent open-weight
coding and agentic workloads when the serving infrastructure is available, and allow for
China-aligned censorship on sensitive topics.
Context window & long-context behaviour
Kimi K2 has a 128K-token context window per the model card. Coverage is partial: treat 128K as
the architectural ceiling, not a verified quality guarantee at full length, and budget KV-cache
memory on the serving cluster.
Kimi-K2-Instruct ships a chat template in its tokenizer configuration. Use it instead of
hand-rolling role markers:
messages = [
{"role": "system", "content": "You are a helpful assistant."},
{"role": "user", "content": "Summarize the following..."},
]
prompt = tok.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
Kimi-K2-Base has no chat template. It is a plain text-completion checkpoint and should not be
prompted as a chat assistant.
Language coverage
Kimi K2 is multilingual with particularly strong English and Chinese. Per-language depth varies
and is not exhaustively documented, so coverage is partial. Validate quality on the target
languages before production use. China-aligned content filtering can affect coverage of sensitive
topics in any language.
Native tool calling and agentic behaviour are a core design goal of Kimi K2. The model is trained
and benchmarked for agentic tasks, and this is its headline differentiator. Tool calling is
exposed through the OpenAI/Anthropic-compatible function-calling API on vLLM and SGLang, so
existing tool routers and agent frameworks work with little adaptation. Treat all tool and
retrieved content as untrusted and add prompt-injection defences (see Implement).
How this scores
The ownership factor this domain covers, drawn from the one entry record.
3
ReliabilityIs it reliable and good enough for the job?
ModerateRunnable but heavy: dependable self-hosting of the trillion-parameter model is a multi-node exercise, and the safety tuning is light with no guard model.
How this scores (AOI sub-dimensions)
Operational3/5how practical it is to run, serve and maintain in productionServing support is good (vLLM, SGLang, KTransformers, TensorRT-LLM) with an OpenAI/Anthropic-compatible API.
Safety3/5whether misuse risks are evaluated and guardrails are providedThe Instruct variant is safety-tuned and withstands casual jailbreaks.
What this means for adoptionYou substantially use, modify and commercialise the self-hosted Kimi K2 weights under a near-MIT grant. The sole condition is the 100M-MAU / $20M-revenue trigger to display 'Kimi K2' in your UI. The weights run entirely on your own infrastructure, so your data stays yours. Ownership is partial because of transparency: the 15.5T-token corpus and training code are closed and the model carries China-aligned censorship that cannot be inspected. The weights and the hosted service differ: the platform.kimi.ai API trains on your content by default (opt-out only by enterprise agreement, Singapore law), so self-host if data control matters. Confirm the attribution trigger and the per-checkpoint LICENSE before shipping, and budget multi-node infrastructure to run it at all.
Sources
The same evidence records as the entry sheet. Read means the text was verified; unverified means it is known to exist but not yet read.
Model cardread2026-11-19
The Kimi-K2-Instruct model card on the verified moonshotai Hugging Face org describes block-FP8 safetensors, 1T total and 32B active parameters, a 128K context and the Muon optimiser.
Licenceread2026-11-19
The Kimi K2 LICENSE is standard MIT with one added clause, which opens "Our only modification part is that".
Documentationunverified2026-07-25
The moonshotai/Kimi-K2 repository documents the MoE architecture (384 experts, 8+1 selected), the MuonClip optimiser, block-FP8 weights and deployment on vLLM, SGLang, KTransformers and TensorRT-LLM with an OpenAI/Anthropic-compatible API.
Technical_reportread2026-07-25
Moonshot's arXiv technical report "Kimi K2: Open Agentic Intelligence" (2507.20534) describes a 1T-total / 32B-active MoE trained on 15.5T tokens with the MuonClip optimiser (Muon + QK-clip) and reports agentic and coding results (e.g.
Terms of serviceread2026-11-19
The platform.kimi.ai model-use agreement (page dated "Last Updated: July 30th, 2026") says user content is used to improve the services, with opt-out available only via an enterprise or separate written agreement.
Privacy Policyread2026-11-19
The platform.kimi.ai privacy policy says the hosted service trains on user prompts, audio, images, videos and files by default; the controller is MOONSHOT AI PTE.
Documentationread2026-11-19
The hosted Kimi API Platform notice says that from 2026-10-31 K2 series models are no longer available to newly registered users (existing users unaffected) and Moonshot plans to retire all K2 series models by the end of December 2026.
Third-party analysisunverified2026-07-25
On independent evaluation Kimi K2 is among the strongest open-weight models on agentic and coding benchmarks (e.g.
Third-party analysisunverified2026-07-25
Independent analysis notes Kimi K2, like other China-based open-weight models, applies China-aligned content filtering on politically sensitive topics.
Third-party analysisunverified2026-07-25
Kimi K2 Instruct is safety-tuned but with lighter alignment coverage than Western frontier labs and no companion guard model.
Third-party analysisunverified2026-07-25
No public EU AI Act training-content summary, copyright policy, or provider documentation package is published for Kimi K2, and the training corpus is not released.
Third-party analysisunverified2026-07-25
Kimi K2 is served across vLLM, SGLang, KTransformers and TensorRT-LLM, but its 1T-parameter scale requires multi-GPU / multi-node infrastructure even in block-FP8.