Der Shop ist online

Wir bauen Workstation AI-Maschinen für Websites, KI-Agenten und LLM-Training – online konfigurieren und kaufen.

Händlerpreise erhalten
KI-Produkte
KI-LabsOpenAI-AgentenClaude-AgentenGrok-BotWorkstation AI-natives CRMMarketingAlle Produkte
KI-Lösungen
Angewandte KIKI-AgentenGenerative KIKI-AutomatisierungKI-WeiterbildungKI-Pakete für KMUEnterprise AI LabKI-WorkstationsNVIDIA-Hardware & -SoftwareNVIDIA Accelerated ComputingPrivate KIGPU-Private-CloudGPU-ClusterEdge-KIKI nach Branche
KI-Services
PlattformmodernisierungDigital EngineeringDatengrundlagen & KIAutonomer BetriebKI-BeratungDevOps-AutomatisierungCybersicherheitSoftwareentwicklungAgentenentwicklungMLOps-EinrichtungGPU-LeasingGPU-Private-Cloud
Über uns
UnternehmenUnsere GeschichtePartnerKundengeschichten
KI-Rechenzentren
KI-RechenzentrenGPU-Private-CloudGPU-ClusterGPU-Leasing
KI-EinblickeJev-ReiheBenchmark LabModellauswahl
Technische Hinweise
Dokumentation
Linux-BetriebssystemeUbuntuRed Hat Enterprise LinuxSUSE LinuxAlpine LinuxDebian, Rocky & AlmaLinux
KI-Cloud-PlattformenAWS & Amazon BedrockGoogle Cloud & Vertex AIMicrosoft Azure & AI FoundryGPU-Clouds & Marktplätze
Private KI-RechenzentrenFür AnwaltskanzleienFür Steuer- und WirtschaftsprüfungskanzleienOn-Premises-PlattformenVollständig abgesicherte Cloud
ToolsRing PromoterJobshout
News
Blog
KontaktAnmelden

KI-Workstations, Multi-Agenten-KI-Software, GPU-Infrastruktur und intelligente Agentenlösungen für moderne Unternehmen.

Kontakt

KI-Lösungen

Angewandte KIKI-AgentenGenerative KIKI-AutomatisierungKI-WeiterbildungKI-Pakete für KMUEnterprise AI LabKI nach Branche

Produkte

Alle ProdukteWorkstation AI-natives CRM & ERPMarketingOpenAI-AgentenRing PromoterJobshout

Unternehmen

Über unsUnternehmenKI-RechenzentrenWarum WorkstationPartnerKundengeschichtenPreiseKontaktTenthMatrix

Ressourcen

KI-EinblickeTechnische HinweiseDokumentationLinux-BetriebssystemeKI-Cloud-PlattformenPrivate KI-RechenzentrenBlogSucheSitemap
Büro Belgien
Workstation SRL, Rue Vanderkindere 34, 1180 Uccle, BrusselsBE 0751.518.683Mo – Fr: 9:00 – 18:00 Uhr MEZ
+32 492 45 67 46
Hauptsitz Großbritannien
77-79 Marlowes, Hemel Hempstead HP1 1LFAnfahrt – Ausfahrt 20 der M25 (Outer London)Handelsregisternr.: 11641870Mo – Fr: 9:00 – 18:00 Uhr GMT
+44 7515 356 146
Büro Indien
#159 Sector 9, Pocket 1, DDA Flats, 110077 Dwarka, New Delhi
+91 98881 98841
Büro Taiwan
3F, No. 2-4, Aisi Rd., Ren'ai Dist., Keelung City 200, Taiwan基隆市仁愛區愛四路2之4號3樓Mo – Fr: 9:00 – 18:00 Uhr Taiwan-Zeit (GMT+8)
+886 926 892 701+32 492 45 67 46
Büro Estland
Aleksander Puškini tn 30-25, 20306 Narva, EstoniaMo - Fr: 9:00 - 18:00 Uhr estnischer Zeit (EET)
+32 492 45 67 46

© 2026 Workstation AI. Alle Rechte vorbehalten.

DatenschutzDatenlöschungCookiesNutzungsbedingungenWebsite-Sitemap

NVIDIA, das NVIDIA-Logo, RTX, RTX PRO, Blackwell, CUDA, NVLink, MGX, Omniverse, Spectrum-X, Quantum, BlueField, DGX und HGX sind Marken und/oder eingetragene Marken der NVIDIA Corporation in den USA und anderen Ländern. AMD, das AMD-Arrow-Logo, EPYC, Instinct, Ryzen und Threadripper sind Marken von Advanced Micro Devices, Inc. Dell und PowerEdge sind Marken der Dell Inc. oder ihrer Tochtergesellschaften. Intel und Xeon sind Marken der Intel Corporation oder ihrer Tochtergesellschaften. AMAX, AceleMax und LiquidMax sind Marken der AMAX Engineering Corporation. Alle anderen Produktnamen, Logos und Marken sind Eigentum ihrer jeweiligen Inhaber. Wir verwenden sie ausschließlich, um die Produkte zu kennzeichnen, die wir beschreiben, liefern oder unterstützen; ihre Verwendung bedeutet nicht, dass ihre Inhaber uns empfehlen.

Startseite / Technische Hinweise / Technology
AIMLOpsOpen Source

Kimi K3 & Open Weights: Frontier Agents Without Locked-In APIs

Workstation deep dive: Kimi K3 specs and license caveats, serving reality, MCP 2026-07-28, expanded open-model radar, hybrid routers, and production checklist

5. August 2026Technology6 min read

Moonshot AI’s Kimi K3 did not invent open weights — Llama, DeepSeek, Qwen, and others already proved the category. What changed in late July 2026 is that an open-weight model at ~2.8 trillion parameters, with a 1M-token context and day-0 production serving, is now competing on agent workloads that used to be reserved for Anthropic and OpenAI. This Workstation guide is for developers and technology leaders who need to decide: self-host, managed API, or hybrid — and which other open models belong on the radar.

Kimi K3 open weights Workstation guide

Companion: Kimi K3 & open-weights — business summary. Related: Claude Opus 5 on AWS, run LLMs on Kubernetes, AI SME packages.

1. What Moonshot shipped

According to Moonshot’s GitHub and technical materials (weights public ~27 Jul 2026):

  • Kimi K3 — open-weight, native multimodal agentic model; ~2.8T total parameters (MoE).
  • Architecture highlights: Kimi Delta Attention (KDA), Attention Residuals (AttnRes), Stable LatentMoE; sparse expert activation (order of 16 of 896 experts / ~100B-class active params per token — confirm on the live model card).
  • 1M-token context; vision + tool use + long-horizon coding/knowledge work.
  • Weights: Hugging Face moonshotai/Kimi-K3 (commonly distributed in MXFP4; footprint on the order of ~1.4TB).
  • Inference engines day-0: vLLM, SGLang, TokenSpeed; hosted API at platform.kimi.ai.

License: Kimi K3 License — not a free-for-all SaaS permit. Enterprises and Model-as-a-Service providers should legal-review before productising (VentureBeat and independent write-ups flagged commercial thresholds). Open weights ≠ open for every business model.

2. Can it keep up with Anthropic and OpenAI?

Short answer for builders: on several agent-relevant benches, yes — close enough that the default should no longer be “closed only.”

Community and vendor evaluations circulating after launch placed K3 in the same band as leading proprietary models on SWE-bench Verified and LiveCodeBench-style suites, with tool-use competitive against Claude Sonnet and GPT-4o-class baselines. That is the milestone: open weights clearing the bar on agent tasks, not just MMLU trivia.

Caveats Workstation insists on:

  • Benchmarks are harness-sensitive — run your golden set.
  • p50 latency wins can hide p95/p99 pain on multi-tool agent chains.
  • Closed labs still lead on some product polish, safety tooling, and support SLAs.
  • Hybrid routing (open for private data + closed for peak-critical) remains the pragmatic architecture.

3. Why infrastructure moved in days, not months

The strategic shift is serving maturity:

  • vLLM published day-0 production guidance: KDA-aware prefix caching, kernels for NVIDIA/AMD, prefill/decode disaggregation, speculative decoding (DSpark draft models), tool calling, structured output.
  • Cloud inference engines and hyperscalers (including AWS SageMaker HyperPod / EKS recipes) published deploy paths the same week weights appeared.
  • Result: choosing self-host vs managed is increasingly a config and FinOps decision, not a quarter-long research project — if you already have GPU capacity and MLOps muscle.

Open-weight stack: weights, serve, govern, agents

4. Hardware reality (do not skip)

K3 is not a laptop model. Moonshot-oriented guidance points at datacenter-scale deployments (often 16+ high-end GPUs as a minimum credible path in published vLLM notes; preferred configs can run to 64+ accelerators). Weights alone are terabyte-class.

SME / mid-market path: run smaller open MoEs on Workstation AI boxes; call K3 via managed API when you need that IQ; keep RAG corpora and agents private.

Enterprise path: GPU cluster + vLLM/SGLang + GitOps + evals — or buy managed capacity and invest engineering in the agent layer instead of kernels.

Video: practical vLLM cloud serve walkthrough — useful before you size a K3-class node.

5. Policy and security backdrop

The same window saw NVIDIA and partners push the Open Secure AI Alliance and amplify the Open Weights and American AI Leadership letter (270+ orgs). The argument: open weights are not only an economics story — defenders need inspectable models to red-team agents and software supply chains. Pair openness with evals, guardrails, and rapid patch culture — not wishful “ban open models.”

6. MCP went stateless (why agents care)

The MCP 2026-07-28 specification is the largest protocol revision since launch:

  • Removes initialize handshake and Mcp-Session-Id — requests carry version/capabilities in _meta.
  • Any instance behind a plain load balancer can serve any request (no sticky Redis session tax).
  • Hardened auth alignment with OAuth/OIDC; formal ~12-month deprecation policy.
  • Extensions: MCP Apps (server-rendered UIs), MCP Tasks (durable long-running job handles).

For Workstation multi-agent stacks, stateless MCP means tool fleets scale like normal HTTP microservices — the missing piece when open models finally get agent-grade IQ.

7. The production gap (still the real story)

Industry surveys (including Mozilla’s open-source AI reporting) keep repeating a pattern: developers try open weights at high rates, but a smaller share reach production than closed-API teams — and the gap often widens with company size. Closed vendors sell a paved road; open models still require you to own serving, scaling, observability, and security. Kimi K3 raises the ceiling; it does not delete the ops half of the job.

8. Models on our radar (expanded)

Beyond K3, Workstation is watching this open / semi-open set for agentic coding and private AI stacks (verify licenses and benches before procurement):

Model Why it matters Fit
Kimi K3 (Moonshot)2.8T MoE, 1M ctx, frontier open agent codingCluster / managed API
Laguna S 2.1 (Poolside)118B-A8B MoE, 1M ctx, strong Terminal/SWE benches, OpenMDWSelf-host SWE agents
Solar Open 2 (Upstage)250B-A15B, 1M ctx, agentic office/coding, KR sovereign angle4–8× H200 class
DeepSeek-V4 familyContinued open MoE pressure on reasoning/coding price-performanceCost-sensitive agents
Qwen 3.x / Max (Alibaba)Broad multilingual + tool ecosystem; distillations for workstationsDefault open stack
Llama 4 class (Meta)Permissive ecosystem, huge fine-tune/community toolingFine-tunes / RAG
Nemotron 3 (NVIDIA)Open weights aligned with NVIDIA serving/security storyGPU-native estates
GLM-4.x / 5 (Zhipu)Strong agent/tooling lineage; watch license + hosting regionTool-heavy agents
KAT-Coder-V2.5 (Kwaipilot)~35B-A3B MoE coding specialist; Apache 2.0; SWE-bench sizedWorkstation SWE
Mistral Large / MagistralEU-friendly commercial open options; strong tooling storyEU private AI
Gemma 3 (Google)Efficient open models for edge and on-deviceEdge / Mac Silicon
Phi-4 class (Microsoft)Small, capable reasoners for constrained boxesSME workstations
Mage-Flow (Microsoft)Compact ~4B text-to-image / edit; MIT; rival larger diffusion stacksLocal creative
Inflect v2 (Owen Song)Tiny local English TTS (Nano/Micro); Apache 2.0Offline voice
Claude Opus 5 / Fable 5 (closed)Still the quality ceiling for many coding/agent paths on BedrockHybrid peak path
GPT-5.6 Sol/Terra/Luna (closed)Tiered closed inference for router designsManaged failover

9. Practical patterns Workstation recommends

Private RAG + tools on open weights (Qwen / Laguna / Solar / DeepSeek)
        │
        ├─ Lead planner on K3 API or Opus 5 (Bedrock) when stuck
        ├─ Builders on mid open MoE self-hosted
        ├─ Review Bot + human gate
        └─ GitOps promote
  1. Write an ADR: default open model + failover closed model.
  2. Golden eval set (50–100 tasks) before flipping production.
  3. Measure p95/p99 on multi-tool agent chains — not median TTFT alone.
  4. Prompt-cache carefully (a UUID at the top of a prompt can nuke hit rates).
  5. Upgrade MCP servers to 2026-07-28 before you scale horizontally.
  6. Size GPUs to the model you will actually serve — not the blog headline.

10. Closing

Kimi K3 is a proof point: open weights now contest frontier agent work, and the serving ecosystem (vLLM, SGLang, cloud recipes) is keeping pace in days. Policy groups are arguing that open models are part of cyber defence, not only cost control. For businesses, the winning move is a hybrid router on Workstation-grade hardware and Multi Agentic Software — open where privacy and cost dominate, closed where quality SLAs demand it, with MCP, evals, and GitOps so “we tried an open model” becomes “we ship on open models.”

Published by Workstation. Specs and benches move weekly — re-check model cards, licenses, and region policy before you buy silicon.

Diesen Artikel teilen

Mehr zu Technology

Laptops für KI-Entwickler: LLMs lokal unter Linux mit GPUs mit 16 GB und 24 GB betreiben (Leitfaden 2026)

Laptops für KI-Entwickler: LLMs lokal unter Linux mit GPUs mit 16 GB und 24 GB betreiben (Leitfaden 2026)

Was auf eine RTX 5080 oder RTX 5090 Laptop GPU passt, die Speicherrechnung, die lohnenden offenen Modelle, die Einrichtung von Ubuntu 26.04 mit dem NVIDIA-Stack, Ollama, llama.cpp und vLLM, Coding-Assistenten, LoRA-Fine-Tuning, Sicherheit und wann Arbeit auf einen Server gehört

Weiterlesen
ML- und LLM-Systeme testen, deren Korrektheit sich nicht beweisen lässt: ein Python-Evidenz-Kit

ML- und LLM-Systeme testen, deren Korrektheit sich nicht beweisen lässt: ein Python-Evidenz-Kit

Evidenz statt Beweise, in etwa 400 Zeilen Python mit echter Ausgabe: Slice-Metriken mit Bootstrap-Intervallen und einem Pass/Fail-Vertrag, Property- und metamorphe Tests mit Hypothesis, Kalibrierung und Conformal Prediction, Pydantic-Verträge und Regressionsevaluierungen für LLM-Ausgaben, Drift-Prüfungen, ein CI-Gate und die exakte Verifikation eines kleinen neuronalen Netzes

Weiterlesen
TorchCurves in der Praxis: lernbare B-Spline- und Legendre-Kurven in PyTorch

TorchCurves in der Praxis: lernbare B-Spline- und Legendre-Kurven in PyTorch

torchcurves installieren, eine Kurve anpassen und dimensionieren, B-Spline oder Legendre und die Eingabeabbildung wählen, eine monotone Kalibrierungskurve und ein additives Modell auf tabellarischen Daten bauen, mit SciPy, scikit-learn, einem KAN und einem MLP vergleichen, Koeffizienten nach JSON exportieren, testen und in der Produktion betreiben, mit echten Ausgaben

Weiterlesen