Model Releases
We've published new research on how we post-train models for accurate search-augmented answers. Our SFT + RL pipeline improves search, citat…
We've published new research on how we post-train models for accurate search-augmented answers. Our SFT + RL pipeline improves search, citation quality, instruction following, and efficiency. With Qwe
We've published new research on how we post-train models for accurate search-augmented answers. Our SFT + RL pipeline improves search, citation quality, instruction following, and efficiency. With Qwen models, we match or beat GPT models on factuality at a lower cost.
Source: Perplexity (X) | 2026-04-22