Support · Kimi K3
Will it last?
Ownership levelPartialnone·limited·partial·substantial·fullAnalytical input D ยท 53.2/100
This page is a projection of the one entry record, the Transparency factor that Support 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
Common problems & fixes
Recurring pitfalls for a 2.8T-parameter MoE are multi-node out-of-memory errors, sharding misconfiguration, MXFP4/MXFP8 kernel-to-engine version mismatches, and checksum or revision drift on download. Pin the revision, verify checksums, and match engine and kernel versions to the card's documented stack. The list is not exhaustive.
Versions, changelog & cadence
Kimi K3 ships as a single checkpoint on the verified moonshotai org (announced around 2026-07-17, open weights around 2026-07-27). There is no Base/Instruct/Thinking split. The Hugging Face revision hash is the changelog anchor; pin it.
A post-trained successor, Kimi K3.1 Pro (Hugging Face, 2026-11-02, its own licence file), is a separate release. Moonshot's hosted platform calls K3 the previous-generation flagship.
Support is the general Moonshot presence: the moonshotai Hugging Face org (model discussions) and the MoonshotAI/Kimi-K3 GitHub repository (issues and the technical report). There is no dedicated support SLA.
Tracked known issues
The model card and technical report show no built-in safety with documented offensive-cyber willingness, undisclosed training data, code and compute, and the operational burden of a 2.8T-parameter model with no small variant. No published issue tracker covers these, so they belong in the deployer's own control stack.
How this scores
The ownership factor this domain covers, drawn from the one entry record.
2
TransparencyDo you know what it is: weights, training, behaviour, and legible terms?
ModerateThe weights can be inspected and a detailed technical report is available, but the training corpus, the training code and the training-token and FLOP figures are closed, so how the model was made cannot be seen or reproduced.
How this scores (AOI sub-dimensions)
Provenance3/5how well we can trace and verify what went into the modelThe weights are distributed from the verified moonshotai org on Hugging Face as native MXFP4/MXFP8 weights (not pickle), with a clear canonical source and no malicious-mirror incident on record (checklist ~4/8).
Governance3/5how accountable and well-documented the publisher isMoonshot AI is an active, accountable publisher with a detailed technical report and a verified hub presence.
What this means for adoptionOwnership is partial. The self-hosted Kimi K3 weights can be used, modified and commercialised, and they run entirely on the deployer's own infrastructure, so data stays with the deployer. The custom, non-OSI licence adds a $20M Model-as-a-Service separate-agreement gate on top of the branding threshold. The training corpus, code and compute are closed. Most consequentially, the model ships with no safety tuning or guard while its own report shows it performing offensive-cyber tasks that frontier labs refuse. Self-host behind a full external control stack, avoid the train-by-default hosted API where data control matters, and confirm the LICENSE and the Model-as-a-Service threshold before commercial deployment.
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 K3 model card on the verified moonshotai Hugging Face org lists a 2.8T total / 104B active MoE (16 of 896 experts), 1M context, MXFP4 weights / MXFP8 activations, native text+image+video, and serving on vLLM/SGLang/TokenSpeed.
Licenceread2026-11-19
The Kimi K3 License grants use, copying, modification, merging, publication, distribution, sublicensing and sale.
Technical_reportread2026-11-19
Moonshot's technical report "Kimi K3: Open Frontier Intelligence" (GitHub PDF) describes a 2.8T/104B MoE with Kimi Delta Attention and Per-Head Muon, and a context curriculum of 8K to 64K (pretrain) and 256K to 1M (cooldown).
Terms of serviceread2026-11-19
The platform.kimi.ai model-use agreement, dated "Last Updated: July 30th, 2026", allows user Content to be used to develop and improve the Services, with opt-out only via an enterprise arrangement on request.
Privacy Policyread2026-11-19
The platform.kimi.ai privacy policy states that the hosted service trains on user prompts, audio, images, videos and files by default ("helps us optimize our models").
Documentationread2026-11-19
The hosted Kimi API Platform model list calls kimi-k3.1-pro "the first model of the new K3.1 series and Kimi's most capable flagship to date" and kimi-k3 "Kimi's previous-generation flagship"; both are live.
Third-party analysisread2026-07-28
Tom's Hardware (reputable coverage) reports on the 2.8T Kimi K3 release.
Third-party analysisunverified2026-07-28
No public EU AI Act training-content summary, copyright policy or provider documentation package exists for Kimi K3, and the training corpus, training code and training-compute figures are not released.