# Tagged classifier

> Announcements and real-world deployments of AI, checked against their sources and scored: open-weights releases, decision models in the wild, and what is actually in use.

- [REAX 1.1: a 27B decision model fine-tuned from pplx-decider](https://shipped.anth.us/decisions/decisionmodels-reax-1-1/index.md): Apache-2.0 weights built on REAX 1, itself a fine-tune of pplx-decider-v1.2. The card reports 71.27 on its own SN92 scoring. (2026-10-21)
- [Decision4Rec releases Kev models that rerank recommender candidates](https://shipped.anth.us/decisions/polignano-decision4rec-kev/index.md): Apache-2.0 adapters from 0.8B to 27B that read a pool of 30 candidates and output a probability for each. Development results only. (2026-10-19)
- [AccuKnox Inc. releases Guard 1.1, an Apache-2.0 guardrail for agent inputs and prompt injection](https://shipped.anth.us/decisions/accuknox-guard-1-1/index.md): A 4.5B guard fine-tuned from a Qwen3.5-4B base that returns a block probability. The card reports 0.802 on a 27,416-item safety set. (2026-10-16)
- [Maincode releases Matilda Decision v1, a 27B multimodal decision model](https://shipped.anth.us/decisions/maincode-matilda-decision-v1/index.md): Apache-2.0 weights on Qwen3.8-27B. The maker's own run scores 65.38 on the public Jev Decision Index 0.3 suite. (2026-10-15)
- [decisionmodels releases s1-llm-auto-router, a 183 MB CPU router under MIT](https://shipped.anth.us/decisions/decisionmodels-s1-llm-auto-router/index.md): A pruned mmBERT-small that answers seven routing questions in one CPU pass. The card reports 0.898 category accuracy on its held-out set. (2026-10-15)
- [StartLux releases Decision v1.1, from 0.8B to 27B, with non-commercial weights](https://shipped.anth.us/decisions/startlux-decision-v1-1/index.md): StartLux-Decision-v1.1 ships Instant and AutoThink variants. Weights are CC BY-NC 4.0; commercial use needs a separate licence. (2026-10-11)
- [Microsoft releases Decision-1, a 9B model that scores fixed answer options](https://shipped.anth.us/decisions/microsoft-decision-1/index.md): Microsoft-Decision-1 returns a probability per answer option instead of prose. Hosted on Foundry and OpenRouter; no licence stated. (2026-10-09)
- [Amazon publishes Strands Decider 2B, a small decision model for agent workflows](https://shipped.anth.us/decisions/amazon-strands-decider-2b/index.md): A 2B LoRA on Qwen3.5-2B-Base, Apache-2.0, with a /v1/systemone server. The card reports 0.779 on 231 public JevBench tasks. (2026-10-09)
- [OpenAI opens its Decisions API in public beta, built on GPT-6 Luna](https://shipped.anth.us/decisions/openai-decisions-api/index.md): A /v1/decisions endpoint returns typed predicate, choice and score answers. Public beta, $0.10 per million input tokens, GA promised in weeks. (2026-10-06)
- [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)

Page 1 of 3 · [next](https://shipped.anth.us/tags/classifier/page/2/index.md)