AI Agent StoreNace.AI · Local decision checkpoint
Drex v1.5 · local checkpoint
Run the downloadable 9B Drex model on your own hardware. It uses the same fixed-answer format as the online service, but has separate hardware, license and setup requirements.
First reviewed 2026-10-11 · Checked 2026-10-15. The launch date has not been confirmed.
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 checkpoint ID: nace-ai/drex-v1.5. The model uses MiMo-V2.6-Distill-Qwen-9B and a pointer head. Python, the provider’s llama.cpp fork and its Ollama fork support /v1/systemone. Each question is scored in one pass; no text is generated.
Weights use the modified Nace.AI Open RAIL-M license, while runtime code is Apache 2.0. The weight license restricts competing products and generally requires a separate license for non-research/non-personal use above $1 million prior-year gross revenue or total funding. Read the exact license and its other use restrictions before deployment.
The documented runner default is 16,384 tokens, with an explicitly enabled maximum of 131,072. In llama-server, context is split across slots: the example -np 2 / -c 32768 gives 16,384 per slot. Native API row and question limits do not establish local runner limits.
Output: Yes/no probabilities, multiple-choice probabilities and ordered rating scores. No generated explanation. Compare decision models, limits and input costs ↗
No hosted token rate or measured operating cost is assigned. The provider describes about 18 GB BF16 weights or 9.5 GB Q8_0, before runtime and context overhead. These are artifact estimates, not measured peak memory or latency.
Provider release and specifications ↗
Check availability and plans ↗
More setup and licensing details ↗
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 ↗