# Decisions in the wild — page 2

> Decision models used in real deployments.

- [Torchcast AI releases Torchcast Decision 27B, built on StartLux weights](https://shipped.anth.us/decisions/torchcast-decision-27b/index.md): A LoRA fine-tune of StartLux-Decision-27B that speaks the /v1/systemone API. Weights are CC BY-NC 4.0; card reports 65.00 on Index 0.2.1. (2026-10-03)
- [SberAI releases FRIDA-Decisions, a Russian and English encoder for structured decisions](https://shipped.anth.us/decisions/sberai-frida-decisions/index.md): A 0.8B MIT-licensed encoder that picks, rates, answers yes/no or ranks in one pass. The card reports 0.893 on 735 razvilka items. (2026-10-02)
- [H2O.ai releases H2O Lightning 4B, an Apache-2.0 decision model with GGUF builds](https://shipped.anth.us/decisions/h2oai-h2o-lightning-4b/index.md): Built on Qwen3.5-4B, with GGUF files from 1.43 GB for llama.cpp. The card reports 88.74% on 231 public JevBench text questions. (2026-10-02)
- [Perplexity releases pplx-decider, an Apache-2.0 27B decision model](https://shipped.anth.us/decisions/perplexity-pplx-decider/index.md): pplx-decider-v1 and v1.1 are open-weight 27B models that return probabilities for fixed answers. Built on Qwen3.8-27B. (2026-10-01)
- [vLLM Semantic Router team releases Decision 2.0 open decision models](https://shipped.anth.us/decisions/vllm-decision-2-0/index.md): Decision 2.0 is an Apache-2.0 family from vllm-sr. The 27B Vega card reports 74.0 on JevArena. Lux is a 9B sibling. (2026-09-29)
- [An independent developer releases Jev-Style 2B Decision v3, a 2B model with a free hosted API](https://shipped.anth.us/decisions/chaoliangunsw-jev-style-2b-decision-v3/index.md): Apache-2.0 fine-tune of Qwen3.5-2B. The card reports 73.6% on public JevBench items against Jev 1.13 at 86.6%. (2026-09-27)
- [Shanghai AI Laboratory releases Intern-Decision, multimodal decision models from 0.8B to 4B](https://shipped.anth.us/decisions/internlm-intern-decision/index.md): Intern-Decision-4B is fine-tuned from Qwen3.5-4B and answers up to 16 questions per pass, with image input. Apache-2.0. (2026-09-26)
- [distil labs releases an invoice-decision model that approves or holds vendor invoices](https://shipped.anth.us/decisions/distil-labs-qwen3-5-4b-invoice-decision/index.md): A 4B Qwen3.5 fine-tune returning one of five decisions. The card reports 98 of 100 correct on synthetic test invoices. (2026-09-21)
- [Jared Palmer releases Kev-9B, an Apache-2.0 LoRA decision model on Qwen3.5-9B](https://shipped.anth.us/decisions/jaredpalmer-kev-9b/index.md): An independent adapter that implements TypeSafe's System One API. The card reports 0.852 on a locked out-of-domain test. (2026-09-20)
- [Convai Innovations releases laya-typed-decisions, a 0.4B ModernBERT decision model](https://shipped.anth.us/decisions/convai-laya-typed-decisions/index.md): An Apache-2.0 encoder fine-tuned on four synthetic workflows. The card reports 0.766 accuracy and warns its probabilities are over-confident. (2026-09-18)

Page 2 of 3 · [previous](https://shipped.anth.us/decisions/index.md) · [next](https://shipped.anth.us/decisions/page/3/index.md)