AI Agent StoreBaseten / Base Labs · Download verified
Mammoth-1 · download
Run fixed-choice classification and scoring on your own hardware, with text or images.
Released 2026-10-21 · First reviewed 2026-10-23 · Checked 2026-11-03.
We still need comparable independent tests. You can compare the confirmed features and limits below, but we cannot yet give this version an Agent Fit or Practical Value rating. Test it on records with known answers, including cases that should go to a person.
What you can use today
You download the model files and supply the hardware and software to run them. Capacity is the amount of information a request can contain; settings and available memory can limit what fits.
Exact download ID: baselabs/mammoth-1. Supply text or images and the choices you want the model to compare. Your application must carry out any action you choose from its answer.
MIT-labelled 27B weights and a pinned CUDA serving setup. The vision encoder is frozen. Check memory for all weights plus request data, and validate confidence on your own examples.
The model card documents 16,384 maximum sequence length. Check the exact service if you use a hosted copy; a local limit does not guarantee that service’s limits.
Output: Yes/no probabilities, choices and ordered scores; no written explanation. Compare decision models, limits and input costs ↗
The hosted launch promotion does not pay for your local hardware. No measured local operating cost or independent comparison is established.
Provider release and specifications ↗
Check availability and plans ↗
Test it before switching
Give each model the same examples, tools and definition of success. Record correct outcomes, response time, retries and total cost. Include the time and hardware needed to run it.
Compare caching and long-context costs ↗