AI Briefing#decision-models
The latest AI and developer news about #decision-models, with the original source and a short summary.
The latest AI and developer news about #decision-models, with the original source and a short summary.
Decider v1.3 achieves 91.6% average accuracy on 63 classification tasks, outperforming previous production classifiers by 6.54 percentage points.
The multimodal model processes text, images, audio, video, and PDFs to return Boolean probabilities, choices, and scores for policy checks and ranking.
The new interface supports choice, score, and noul primitives with a unified API across transformers and llama.cpp.
Decision models now handle one in four gateway requests at a fraction of the cost, while Anthropic becomes the first choice for nearly twice as many new teams as OpenAI.
The open-source family of four models uses non-autoregressive decoding to achieve 67.52 on Decision Index 0.4.1 and rank top 3 on OpenRouter.
The 50-page report details architecture, training, and evaluation for Drex models, which achieve 93.4% accuracy on long-context benchmarks.
A hybrid system routing uncertain cases to Claude Opus 5.5 achieves 95.3% accuracy on hard identity resolution tasks at $2.94 per 1,000, compared to $5.39 for Opus alone.
The free Python package ports models like Cloudflare's Clef and Perplexity's pplx-decider to MLX, achieving sub-20ms latency on small encoders on an M5 Pro and 96.4% accuracy on TREC after fine-tuning.
The 9B open-weights model achieves a 65 score on the AA Decision Index, matching d1-ultra while costing approximately $23 less per 1M tasks.
Mammoth-1 (27B) and Elephant-1 (9B) offer 16k-token context and return choices, probabilities, and scores for text classification and routing.
AutoJev-27B, d1-ultra, and decider-12b tie for the top spot with a score of 60 in the new benchmark for single-pass decision models.
The new multimodal decision models, Decider V1.2 27B and Decider Flash V1.2 4B, classify content and route requests via typed questions.
The new 8B and 30B models integrate a Decision aLoRA, Guardian aLoRA, and the fastest ASR model on the OpenASR Leaderboard while maintaining base-model VRAM usage.
The multimodal model returns structured probabilities for classification and routing tasks rather than generating text.
The update introduces a playground tab for testing decision models like Kev and OpenJev with probability bars.
Integrating the hosted Jev decision model with Qwen 3.8 27B reduced execution API costs by 26.6% while maintaining 100% task accuracy on 360 SOP-Bench cases.
The new Apache-2.0 model family, ranging from 0.6B to 27B parameters, achieves state-of-the-art scores on both text and vision boards of the Jev Decision Index.
The new model, available in Microsoft Foundry, delivers calibrated probability scores for routing and classification tasks at 35 times the speed of GPT-6 Sol.
The d1 model returns structured answers with probabilities for classification and routing without generating text tokens.
The new d1-3B model achieves a score of 48.57 on the Decision Index 0.2.1, outperforming larger models like Decider 35B-A3B.
Clef models are open-sourced under Apache 2.0 on Hugging Face and hosted on Workers AI, claiming to outperform Jev on the Decision Index.
The 2-billion-parameter model offers sub-150ms latency on local hardware and ranks 3rd of 33 in its class on JevBench accuracy.
The release includes Kev-27B and an updated Kev-9B, offering Apache-2.0 weights compatible with TypeSafe's Jev API for local and cloud deployment.
Jeeves achieves 0.935 on JevBench overall, outperforming the baseline Jev model's 0.866 by integrating reasoning chains and a block-4 diffusion drafter.