diff --git a/crates/xserv-model/src/sampling.rs b/crates/xserv-model/src/sampling.rs
index 762b62e..efa2c4c 100644
--- a/crates/xserv-model/src/sampling.rs
+++ b/crates/xserv-model/src/sampling.rs
@@ -19,6 +19,18 @@ impl Default for SamplingParams {
 /// Uses the last position's logits. Handles both F32 and BF16 dtypes.
 pub fn sample(logits: &Tensor, params: &SamplingParams) -> u32 {
     assert_eq!(logits.ndim(), 2);
+    // Greedy fast path: GPU argmax + 4-byte D2H instead of copying the whole
+    // [seq, vocab] logits to the host and scanning it (~201k bf16/token).
+    // NaN logits lose every `>` comparison in the kernel, matching the
+    // NaN-safe host argmax below.
+    if params.temperature == 0.0
+        && logits.dtype() == DType::BF16
+        && matches!(logits.device(), Device::Cuda(_))
+        && logits.is_contiguous()
+    {
+        let ids = xserv_kernels::argmax_bf16_to_host(logits);
+        return *ids.last().unwrap();
+    }
     let vocab_size = logits.shape()[1];
     let seq_len = logits.shape()[0];
     let logits_cpu = logits.to_device(Device::Cpu);