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.
A 4B decision model fine-tuned from Qwen3.5-4B-Base. A state and typed questions (choice, yes/no, score) go in; a temperature-calibrated probability for every option comes out of one pass, through a self-hosted /v1/systemone-compatible server. Apache-2.0 weights and code.
Decides
choice, score, noul, classify, route
choice, score, noul, classify, route
Architecture
vega
wald
Fine-tuned from
qwen/qwen3.5-0.8b
qwen/qwen3.5-4b-base
License
apache-2.0
apache-2.0
Availability
Open weights
Open weights
Hosted by
—
—
Input price
—
—
Decision accuracy
55.6%
88.3%
Calibration error
0.151
0.045
Valid action rate
—
—
Median latency
—
33 ms
p95 latency
—
168 ms
Evaluation suite
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 vega and wald?
vega is from Nandakishor M and wald from ORG2 AI. Both have open weights you can download and run. Both answer choice, score, noul, classify and route questions. wald reads up to 128K tokens of state, against 72K tokens for vega. vega is the smaller model, at 800M parameters to 4.2B.
Which is more accurate, vega or wald?
They report on different suites — 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, wald 88.3% on JevBench public set (231 items; 204 correct), maker's run with the JevBench CLI, v1.2, effort none; items used as a development scoreboard — so the numbers do not rank them. Test both on your own labelled examples.
Which is cheaper, vega or wald?
vega: Free (open weights). wald: Free (open weights). Open weights cost nothing per call beyond your own hardware.
Can I run vega or wald locally?
Yes, both: systemone pull nandakishorm/vega and systemone pull org2ai/wald download the weights.
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
JevBench public set (231 items; 204 correct), maker's run with the JevBench CLI, v1.2, effort none; items used as a development scoreboard
Latest version
2026.10.12
1.2.0
Variants
adapters, backbone, 4b
Dockerfile, LICENSE, NOTICE, model-files, reference, server
