Open-weight or API? How to decide when the break-even numbers disagree

    This summer brought a run of large open-weight models. Moonshot AI released Kimi K3, a 2.8-trillion-parameter model, on 16 July 2026. Z.ai’s GLM-5.2 (753 billion parameters, reported under an MIT license) and Alibaba’s Qwen 3.8-Max (2.4 trillion) arrived in the same stretch. One report says open-weight models have narrowed the gap with proprietary systems to single digits on most benchmarks that matter to enterprise buyers. So the question gets asked a lot: should we host one ourselves?

    The cost question, and why the answers disagree

    API prices vary widely. One comparison lists DeepSeek V4-Pro at $0.87 per million output tokens and Kimi K3 at $15, a gap of more than 17 times. That alone changes the sums. But published break-even points for self-hosting don’t agree. A developer write-up on Kimi K3 puts the break-even at roughly 2.7 billion output tokens a month. CoreWeave, which sells GPU infrastructure, says cost analyses this year put it near 5 to 10 million tokens a day for a right-sized open model.

    Figure 1 · Two published break-even estimates

    CoreWeave, low end5M
    CoreWeave, high end10M
    Kimi K3 write-upabout 90M
    Break-even for self-hosting, in millions of tokens per day. The Kimi K3 figure is the monthly number divided by 30. The models, hardware and assumptions differ, and CoreWeave has a commercial interest, so compare the spread, not the exact values.

    A spread this wide says the answer depends on the model you pick, the hardware you can keep busy, and what you count as cost. Do your own sums with your own volumes.

    A way to decide

    Figure 2 · Choosing a path

    QuestionLeans APILeans self-host
    VolumeLow or unevenHigh and steady, so the GPUs stay busy
    Data rulesProvider terms are acceptableData must not leave your network
    TeamNo one to run GPUsYou can operate and monitor the stack
    Model needsBest quality on hard tasksA smaller model is good enough
    LicenseNot your concernRead it. Reports on K3’s terms conflict
    Our own rule of thumb, not a published benchmark.

    One practical point: a self-hosted model only saves money when you keep expensive hardware busy around the clock. And licenses differ. Some sources call Kimi K3’s license MIT, others describe a custom license with conditions for hosting providers, so read the actual text before you plan around it.

    It doesn’t have to be either

    A mixed setup is often the sensible one: fast, cheap models for intake and classification, and a larger model, hosted by you or by a provider, for the hard reasoning. That is a routing problem, and we covered the thinking in routing belongs in the harness.

    Sources

    Want help putting this into practice?

    Book a free consultation and tell us about your team or school.

    Book a free consultation