diff --git a/crates/llama-cpp-bindings/src/engine.cc b/crates/llama-cpp-bindings/src/engine.cc
index 30014f4..3c2e980 100644
--- a/crates/llama-cpp-bindings/src/engine.cc
+++ b/crates/llama-cpp-bindings/src/engine.cc
@@ -29,6 +29,7 @@ struct Request {
   std::vector<llama_token> tokens;
   size_t i_batch = -1;
   size_t n_past = 0;
+  size_t n_draft = 0;
 
   int32_t multibyte_pending = 0;
   std::string generated_text;
@@ -213,50 +214,58 @@ class TextInferenceEngineImpl : public TextInferenceEngine {
           continue;
         }
 
-        auto logits = llama_get_logits_ith(ctx, i_batch);
-        auto next_token = std::distance(logits, std::max_element(logits, logits + n_vocab));
+        // FIXME: ensure batching logic always put i_batch - request.n_draft in this batch.
+        for (int k = -request.n_draft; k < 1; ++k) {
+          auto logits = llama_get_logits_ith(ctx, i_batch);
+          auto next_token = std::distance(logits, std::max_element(logits, logits + n_vocab));
 
-        request.n_past += request.tokens.size();
+          request.n_past += request.tokens.size();
 
-        request.tokens.clear();
-        request.tokens.push_back(next_token);
+          request.tokens.clear();
+          request.tokens.push_back(next_token);
 
-        const auto token_str = llama_token_to_piece(ctx, next_token);
-        request.generated_text += token_str;
+          const auto token_str = llama_token_to_piece(ctx, next_token);
+          request.generated_text += token_str;
 
-        // FIXME: Hack for codellama to simplify tabby's implementation.
-        const bool is_eos = next_token == eos_id || token_str == " <EOT>";
+          // FIXME: Hack for codellama to simplify tabby's implementation.
+          const bool is_eos = next_token == eos_id || token_str == " <EOT>";
 
-        if (request.multibyte_pending > 0) {
-          request.multibyte_pending -= token_str.size();
-        } else if (token_str.size() == 1) {
-          const char c = token_str[0];
-          // 2-byte characters: 110xxxxx 10xxxxxx
-          if ((c & 0xE0) == 0xC0) {
-            request.multibyte_pending = 1;
-            // 3-byte characters: 1110xxxx 10xxxxxx 10xxxxxx
-          }
-          else if ((c & 0xF0) == 0xE0) {
-            request.multibyte_pending = 2;
-            // 4-byte characters: 11110xxx 10xxxxxx 10xxxxxx 10xxxxxx
-          } else if ((c & 0xF8) == 0xF0) {
-            request.multibyte_pending = 3;
-          }
-          else {
-            request.multibyte_pending = 0;
-          }
-        }
-
-        if (request.multibyte_pending == 0) {
-          rust::String generated_text;
-          try {
-            generated_text = is_eos ? "" : request.generated_text;
-          } catch (const std::invalid_argument& e) {
-            fprintf(stderr, "%s:%d [%s] - ignoring non utf-8/utf-16 output\n", __FILE__, __LINE__, __func__);
+          if (request.multibyte_pending > 0) {
+            request.multibyte_pending -= token_str.size();
+          } else if (token_str.size() == 1) {
+            const char c = token_str[0];
+            // 2-byte characters: 110xxxxx 10xxxxxx
+            if ((c & 0xE0) == 0xC0) {
+              request.multibyte_pending = 1;
+              // 3-byte characters: 1110xxxx 10xxxxxx 10xxxxxx
+            }
+            else if ((c & 0xF0) == 0xE0) {
+              request.multibyte_pending = 2;
+              // 4-byte characters: 11110xxx 10xxxxxx 10xxxxxx 10xxxxxx
+            } else if ((c & 0xF8) == 0xF0) {
+              request.multibyte_pending = 3;
+            }
+            else {
+              request.multibyte_pending = 0;
+            }
           }
 
-          result.push_back({request.id, generated_text});
-          request.generated_text.clear();
+          if (request.multibyte_pending == 0) {
+            rust::String generated_text;
+            try {
+              generated_text = is_eos ? "" : request.generated_text;
+            } catch (const std::invalid_argument& e) {
+              fprintf(stderr, "%s:%d [%s] - ignoring non utf-8/utf-16 output\n", __FILE__, __LINE__, __func__);
+            }
+
+            result.push_back({request.id, generated_text});
+            request.generated_text.clear();
+          }
+
+          if (k < 0 && next_token != request.tokens[request.tokens.size() + k]) {
+            // FIXME: shift kv cache
+            break;
+          }
         }
       }
     }