Wald-4B

ORG2 AIDecision modelOpen weights

Wald-4B by ORG2 AI is a decision model: it answers typed questions with a probability for each allowed answer instead of generating text.

Pricing

No hosted price is listed. The weights are open: the cost is your own compute.

Specifications

Provider: ORG2 AI
Parameters: 5B
Released: Sep 2026
Open weights: Yes
Question types: Choice (one of N options), Score (ordered rubric), Yes / no
Mechanism: Decoder LM with a decision readout
Backbone: Qwen/Qwen3.5-4B
Jev-compatible request: Yes
Runs on CPU: No
License: apache-2.0
Hugging Face: org2ai/Wald-4B

Benchmarks

Benchmark Heaven JevBench (Capability): 69.3

Self-reported by the publisher

The model's own claims, often on its own sets or a self-run copy of a public suite. Not independently verified and never ranked alongside measured results.

Jev Decision Index public suite (self-run)Self-reportedDecision Index public balanced skill, edition 0.3 kit (release-v3), self-run: 60.0

Compare Wald-4B with…

Wald-4B vs amos-1-large—/1M in Wald-4B vs amos-1-medium—/1M in Wald-4B vs Celeris-1 Decision$0.04/1M in Wald-4B vs ScaleDown Classify-1$0.04/1M in Wald-4B vs Cloudflare clef—/1M in Wald-4B vs Cloudflare clef-flash—/1M in

FAQ

How much does Wald-4B cost?

Wald-4B has no hosted price listed; its weights are open, so the cost is the compute you run it on.

What questions can Wald-4B answer?

Wald-4B answers choice (one of n options), score (ordered rubric), yes / no questions, returning a probability for each allowed answer.

Is Wald-4B well calibrated?

No calibration measurement for Wald-4B is recorded yet.

Is Wald-4B open weight, and can it run on a CPU?

Yes, Wald-4B 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 (Benchmark Heaven JevBench, Hugging Face). Pricing is per 1M tokens (USD); confirm with the provider before production use.