Model2d ago
GPT-5.5 Reclaims Benchmark Lead Despite Higher Costs
A 20 percent API price hike accompanies the release of GPT-5.5. The model regains the top spot on industry benchmarks, though frequent hallucinations persist. OpenAI prioritizes raw performance over reliability here. Practitioners must weigh these marginal accuracy gains against increased operational costs and lingering factual errors.