> Multiple models are close to or exceeding GPT 3.5 now Only in benchmarks as GP said. All the models I have tried either hallucinates like crazy or refuse to

> Multiple models are close to or exceeding GPT 3.5 now Only in benchmarks as GP said. All the models I have tried either hallucinates like crazy or refuse to answer everything. Nowhere close to GPT 3.5.