mmastrac diffusiongemma (vLLM PR 57250)
mmastrac diffusiongemma (vLLM PR 57250) by mmastrac is a decision model: it answers typed questions with a probability for each allowed answer instead of generating text. It scores 36.3 on the Jev Decision Index (rank 95).
Pricing
No hosted price is listed. The weights are open: the cost is your own compute.
Specifications
- Provider
- mmastrac
- Parameters
- 26B
- Released
- Sep 2026
- Open weights
- Yes
- Question types
- Choice (one of N options), Score (ordered rubric), Yes / no
- Mechanism
- Diffusion LM
- Backbone
- google/diffusiongemma-26B-A4B-it
- Jev-compatible request
- No
- Runs on CPU
- No
- License
- —
Benchmarks
- Jev Decision IndexECE 0.202
- 36.3
- Jev Decision Index — VisionECE 0.136
- 46.3
Jev Decision Index: public 32.7, ECE 0.202, Brier 0.535. Lower ECE means its probabilities can be taken at face value.
Speed & latency
Compare mmastrac diffusiongemma (vLLM PR 57250) with…
FAQ
How much does mmastrac diffusiongemma (vLLM PR 57250) cost?
mmastrac diffusiongemma (vLLM PR 57250) has no hosted price listed; its weights are open, so the cost is the compute you run it on.
What questions can mmastrac diffusiongemma (vLLM PR 57250) answer?
mmastrac diffusiongemma (vLLM PR 57250) answers choice (one of n options), score (ordered rubric), yes / no questions, returning a probability for each allowed answer.
Is mmastrac diffusiongemma (vLLM PR 57250) well calibrated?
On the Jev Decision Index its expected calibration error (ECE) is 0.202; below about 0.05, a stated 80% confidence is right roughly 80% of the time.
Is mmastrac diffusiongemma (vLLM PR 57250) open weight, and can it run on a CPU?
Yes, mmastrac diffusiongemma (vLLM PR 57250) has open weights. It needs a GPU to run at useful speed.
Data as of 2026-10-14. Source: Metatext decision-model catalogue, 2026-10-14 (Jev Decision Index). Pricing is per 1M tokens (USD); confirm with the provider before production use.