851
Mystery company accidentally blew $500 million on Claude in a single month — failed to put usage limit on licenses for employees (www.tomshardware.com)
submitted 3 months ago by to c/technology@lemmy.world
168 comments fedilink hide all child comments
you are viewing a single comment's thread
view the rest of the comments
[–] 2 points 3 months ago (5 children)

Depends how big the model is. Smaller models can be dirt cheap.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 5 child comments
    •
  • [–] 1 point 3 months ago* (4 children)

    Think deeper in Copilot is apparently o3-mini, or it used to be until recently. GPT 5.5 (and some other versions) are also an option, so I should try using these next week.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 4 child comments
    •
  • [–] 2 points 3 months ago (3 children)

    If you have a high VRAM GPU (like 16GB or more) you can try running some model locally with LMStudio. You won't get the same level as cloud models but depending on use case it might be good enough.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 3 child comments
    •
  • [–] 1 point 3 months ago (2 children)

    I only have an Radeon RX Vega 56 8GB HBM2, so limited to CPU-only. 64 GB memory there at least. Not upgrading unless hardware dies.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 2 child comments
    •
  • [–] 3 points 3 months ago (1 child)

    RX Vega 56 8GB HBM2, 64GB RAM

    That's poor but not horrible. LMStudio has Vulkan compute backend option that works on almost any card. You can also do partial GPU/RAM split giving you total of 72GB and increased performance vs CPU only.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 1 child comment
    •