From bdc1dda64f80393af9255128dcdf70ea08e3e046 Mon Sep 17 00:00:00 2001
From: Jesse Posner <jesse.posner@gmail.com>
Date: Sun, 15 Feb 2026 22:25:12 -0800
Subject: [PATCH] chat : remove dead thinking code from qwen3_coder_xml

Remove thinking handling code that became unreachable after routing
Step-3.5-Flash to the Nemotron v3 PEG parser. Qwen3-Coder has no
<think> in its template, so the thinking_forced_open logic, preserved
tokens, and grammar prefix were dead paths.
---
 common/chat-parser-xml-toolcall.cpp |  1 -
 common/chat.cpp                     | 11 -----------
 tests/test-chat.cpp                 |  3 +--
 3 files changed, 1 insertion(+), 14 deletions(-)

diff --git a/common/chat-parser-xml-toolcall.cpp b/common/chat-parser-xml-toolcall.cpp
index 56d8bb410a..a80900ff8d 100644
--- a/common/chat-parser-xml-toolcall.cpp
+++ b/common/chat-parser-xml-toolcall.cpp
@@ -279,7 +279,6 @@ void build_grammar_xml_tool_call(common_chat_params & data, const json & tools,
             auto call_end = builder.add_rule("root-call-end", form.last_tool_end ? gbnf_format_literal(*form.last_tool_end) : gbnf_format_literal(form.tool_end));
             auto tool_call_multiple_with_end = builder.add_rule("root-tool-call-multiple-with-end", tool_call_once + " " + tool_call_more + "* " + call_end);
             builder.add_rule("root",
-                std::string(data.thinking_forced_open ? "( \"</think>\" space )? " : "") +
                 (form.scope_start.empty() ? "" : gbnf_format_literal(form.scope_start) + " ") +
                 tool_call_multiple_with_end  + "?" +
                 (form.scope_end.empty() ? "" : " " + gbnf_format_literal(form.scope_end))
diff --git a/common/chat.cpp b/common/chat.cpp
index d87cb682c5..04fe8bc070 100644
--- a/common/chat.cpp
+++ b/common/chat.cpp
@@ -1884,18 +1884,7 @@ static common_chat_params common_chat_params_init_qwen3_coder_xml(const common_c
     data.prompt = apply(tmpl, params);
     data.format = COMMON_CHAT_FORMAT_QWEN3_CODER_XML;
 
-    // Handle thinking tags (e.g. Step-3.5-Flash unconditionally emits <think>)
-    if (string_ends_with(data.prompt, "<think>\n")) {
-        if (!params.enable_thinking) {
-            data.prompt += "</think>";
-        } else {
-            data.thinking_forced_open = true;
-        }
-    }
-
     data.preserved_tokens = {
-        "<think>",
-        "</think>",
         "<tool_call>",
         "</tool_call>",
         "<function=",
diff --git a/tests/test-chat.cpp b/tests/test-chat.cpp
index 1bf6e4d9ec..1bef5b9f44 100644
--- a/tests/test-chat.cpp
+++ b/tests/test-chat.cpp
@@ -3566,14 +3566,13 @@ Hey there!<|im_end|>
             common_chat_templates_inputs inputs;
             inputs.messages = { message_user };
             inputs.tools = { special_function_tool };
-            inputs.enable_thinking = true;
             auto params = common_chat_templates_apply(tmpls.get(), inputs);
             assert_equals(COMMON_CHAT_FORMAT_PEG_CONSTRUCTED, params.format);
             assert_equals(true, params.thinking_forced_open);
             assert_equals(false, params.grammar.empty());
             assert_equals(false, params.parser.empty());
             auto grammar = build_grammar(params.grammar);
-            GGML_ASSERT(grammar && "Failed to build Step-3.5-Flash grammar with thinking_forced_open");
+            GGML_ASSERT(grammar && "Failed to build Step-3.5-Flash grammar");
         }
     }
 }