Deft writing model launches, claiming 86% of queries pass as human
TuhinChakr · x · 2026-08-19
New AI lab Deft released its eponymous writing model, focusing on generating human-like text that evades detection through detailed prompts and style adjustments. The lab claims that in current tests, 86% of user queries are judged as fully human-written. The model is in public beta and performs better in analysis, essays, and creative writing, but is weaker in marketing copy and news generation. Some critics questioned the marketing focus on 'fooling AI detectors' rather than improving writing quality.
More from Models
- Grok Bot Ran a 12-Hour Data Analysis Nonstop, Handling API Rate Limits Itself — Daniel_Farinax · 2026-10-16
- One year of Recursive Language Models: GPT-5-mini beats GPT-5 by 110% on 132k-token tasks, cheaper too — a1zhang · 2026-10-16
- Users Report Claude Models Suddenly More Compliant After Forced Compaction — RealSharpNinja · 2026-10-16
- Prediction: Opus 5.5-level models on laptops by 2027, backed by Qwen 27B data — cephaloform · 2026-10-16
- Users Say Krea 2 Output Degraded Sharply Since Krea 3 Beta Launched — szywis · 2026-10-16
- LLM-as-a-Verifier lifts Nemotron 3.5 Super by 9-10 points on agentic benchmarks — simonguozirui · 2026-10-16