French medical decision models by Bofeng Huang: pass a patient message or clinical note, a question and candidate answers, and get one probability per answer from a single forward pass. Choice, score and noul. A research model, not a medical device.
Nandakishor M's typed-decision model. A frozen Qwen3.5 backbone is read, never trained, and cut off mid-stack; a small trained physics engine settles each choice, score or yes/no question into calibrated probabilities, a conformal set and an abstain flag. 0.8B and 4B.
Decides
choice, score, noul
choice, score, noul, classify, route
Architecture
docto-decision
vega
Fine-tuned from
qwen/qwen3.5-4b
qwen/qwen3.5-0.8b
License
apache-2.0
apache-2.0
Availability
Open weights
Open weights
Hosted by
—
—
Input price
—
—
Decision accuracy
88.9%
55.6%
Calibration error
0.079
0.151
Valid action rate
—
—
Median latency
47 ms
—
p95 latency
—
—
Figures are from each model’s manifest; accuracy and latency are what the publishers report, on their own suites and hardware. Add a third model.
Questions
What is the difference between docto-decision and vega?
docto-decision is from Bofeng Huang and vega from Nandakishor M. Both have open weights you can download and run. Both answer choice, score and noul questions. Only vega answers classify and route. vega is the smaller model, at 800M parameters to 4.7B.
Which is more accurate, docto-decision or vega?
They report on different suites — docto-decision 88.9% on Docto Decision Bench fr v0.1 (12 tasks, mostly silver labels; the maker's own benchmark), vega 55.6% on His 11 October 2026 benchmark, published with its results and its harness (a notebook in vegaml): macro accuracy over 43 public tasks for the 0.8B, 0.717 for the 4B; with JevBench (231 items) and DecisionBench (1,075 rows), 3,841 decisions in all, seed 20261011. On his own set of 2,050 choice decisions over six intent workflows (esci, clinc150, appliances, banking77, when2call, apibank), one choice question each, which he has not published with its question wording, he reports 0.763 (calibration error 0.026) for the 0.8B with its public_intents adapter, 0.389 without the adapter and 0.803 for the 4B with it. He publishes no accuracy for the typed_decisions adapters — so the numbers do not rank them. Test both on your own labelled examples.
Which is cheaper, docto-decision or vega?
docto-decision: Free (open weights). vega: Free (open weights). Open weights cost nothing per call beyond your own hardware.
Can I run docto-decision or vega locally?
Yes, both: systemone pull bofeng-huang/docto-decision and systemone pull nandakishorm/vega download the weights.
Evaluation suite
Docto Decision Bench fr v0.1 (12 tasks, mostly silver labels; the maker's own benchmark)
His 11 October 2026 benchmark, published with its results and its harness (a notebook in vegaml): macro accuracy over 43 public tasks for the 0.8B, 0.717 for the 4B; with JevBench (231 items) and DecisionBench (1,075 rows), 3,841 decisions in all, seed 20261011. On his own set of 2,050 choice decisions over six intent workflows (esci, clinc150, appliances, banking77, when2call, apibank), one choice question each, which he has not published with its question wording, he reports 0.763 (calibration error 0.026) for the 0.8B with its public_intents adapter, 0.389 without the adapter and 0.803 for the 4B with it. He publishes no accuracy for the typed_decisions adapters
