Microsoft Decision-1 vs JEV vs Perplexity and EVERY Decision Model (I Tested 20)
Anass Kartit compares sixteen hosted decision models and four local setups on his own 82-question test. Perplexity’s Decider leads his results; Jev scores 79 out of 82 and Microsoft Decision-1 scores 75. He also examines malicious instructions, reordered choices and differences between local runtimes.
Original by AnassKartitEvaluationIntermediate5 min 51 secPublished
Supply the answers for a decision model to score instead of asking it to write a reply. The video shows a request and response that choose an issue category.
Include hostile instructions in testing, not just ordinary questions. Kartit reports Microsoft Decision-1 obeying an injected instruction despite staying stable when answer choices were reordered.
For a local model, check which software runs it. Kartit reports a fourteen-point accuracy difference between runtimes and a memory-related freeze. His Mac results do not establish how other machines perform.
Worth knowing
Kartit's eighty-two-item comparison is one creator's test, not a general ranking. He reports the scores, latencies and costs, measuring cloud timing from his connection. The local software and memory findings are from his Apple Silicon setup. No benchmark was reproduced here.