server: add minimax-m2 reasoning format override for MiniMax-M2 compatibility

MiniMax-M2 models require the complete <think>...</think> block including tags
to be present in the context for proper reasoning. This mode injects a synthetic
opening <think> tag in the stream while keeping all reasoning tags inline in
message.content, ensuring the model receives the full reasoning block it needs.

Changes:
- Add COMMON_REASONING_FORMAT_MINIMAX_M2 enum value to common_reasoning_format
- Implement minimax-m2 format parsing that bypasses reasoning extraction
- Inject synthetic <think>\n chunk at slot start when minimax-m2 is active
- Track injection state with minimax_reasoning_prefix_injected slot flag
- Prepend <think>\n to generated_text for final response and chat parsing
- Prevent client reasoning_format=auto from overriding server CLI setting
- Add minimax-m2 to CLI help, README.md, and code documentation
- Handle LLAMA_TOKEN_NULL in send_partial_response to skip token recording
- Update process_token to preserve delta_to_send for streaming correctness

Author: ServeurpersoCom

common/arg.cpp

@@ -3442,6 +3442,7 @@ common_params_context common_params_parser_init(common_params & params, llama_ex
         "- none: leaves thoughts unparsed in `message.content`\n"
         "- deepseek: puts thoughts in `message.reasoning_content`\n"
         "- deepseek-legacy: keeps `<think>` tags in `message.content` while also populating `message.reasoning_content`\n"
+        "- minimax-m2: streams a synthetic opening `<think>` and keeps `</think>` tags in `message.content`\n"
         "(default: auto)",
         [](common_params & params, const std::string & value) {
             params.reasoning_format = common_reasoning_format_from_name(value);

common/chat-parser.cpp

@@ -171,7 +171,8 @@ void common_chat_msg_parser::consume_literal(const std::string & literal) {
 bool common_chat_msg_parser::try_parse_reasoning(const std::string & start_think, const std::string & end_think) {
     std::string pending_reasoning_prefix;
 
-    if (syntax_.reasoning_format == COMMON_REASONING_FORMAT_NONE) {
+    if (syntax_.reasoning_format == COMMON_REASONING_FORMAT_NONE ||
+        syntax_.reasoning_format == COMMON_REASONING_FORMAT_MINIMAX_M2) {
         return false;
     }
 

common/chat.cpp

@@ -655,6 +655,7 @@ const char * common_reasoning_format_name(common_reasoning_format format) {
         case COMMON_REASONING_FORMAT_AUTO:     return "auto";
         case COMMON_REASONING_FORMAT_DEEPSEEK: return "deepseek";
         case COMMON_REASONING_FORMAT_DEEPSEEK_LEGACY: return "deepseek-legacy";
+        case COMMON_REASONING_FORMAT_MINIMAX_M2: return "minimax-m2";
         default:
             throw std::runtime_error("Unknown reasoning format");
     }
@@ -669,6 +670,8 @@ common_reasoning_format common_reasoning_format_from_name(const std::string & fo
         return COMMON_REASONING_FORMAT_DEEPSEEK;
     } else if (format == "deepseek-legacy") {
         return COMMON_REASONING_FORMAT_DEEPSEEK_LEGACY;
+    } else if (format == "minimax-m2") {
+        return COMMON_REASONING_FORMAT_MINIMAX_M2;
     }
     throw std::runtime_error("Unknown reasoning format: " + format);
 }
@@ -1790,7 +1793,8 @@ static void common_chat_parse_deepseek_v3_1(common_chat_msg_parser & builder) {
         // </think><｜tool▁calls▁begin｜><｜tool▁call▁begin｜>function<｜tool▁sep｜>NAME\n```json\nJSON\n```<｜tool▁call▁end｜><｜tool▁calls▁end｜>
         common_chat_parse_deepseek_v3_1_content(builder);
     } else {
-        if (builder.syntax().reasoning_format == COMMON_REASONING_FORMAT_NONE) {
+        if (builder.syntax().reasoning_format == COMMON_REASONING_FORMAT_NONE ||
+            builder.syntax().reasoning_format == COMMON_REASONING_FORMAT_MINIMAX_M2) {
           LOG_DBG("%s: reasoning_format none, adding content\n", __func__);
           common_chat_parse_deepseek_v3_1_content(builder);
           return;
@@ -2001,7 +2005,9 @@ static void common_chat_parse_gpt_oss(common_chat_msg_parser & builder) {
 
         if (regex_match(analysis_regex, header)) {
             builder.move_to(header_start_pos);
-            if (builder.syntax().reasoning_format == COMMON_REASONING_FORMAT_NONE || builder.syntax().reasoning_in_content) {
+            if (builder.syntax().reasoning_format == COMMON_REASONING_FORMAT_NONE ||
+                builder.syntax().reasoning_format == COMMON_REASONING_FORMAT_MINIMAX_M2 ||
+                builder.syntax().reasoning_in_content) {
                 builder.add_content(consume_end(true));
             } else {
                 builder.try_parse_reasoning("<|channel|>analysis<|message|>", "<|end|>");

common/common.h

@@ -249,6 +249,7 @@ enum common_reasoning_format {
     COMMON_REASONING_FORMAT_AUTO,            // Same as deepseek, using `message.reasoning_content`
     COMMON_REASONING_FORMAT_DEEPSEEK_LEGACY, // Extract thinking tag contents and return as `message.reasoning_content`, or leave inline in <think> tags in stream mode
     COMMON_REASONING_FORMAT_DEEPSEEK,        // Extract thinking tag contents and return as `message.reasoning_content`, including in streaming deltas.
+    COMMON_REASONING_FORMAT_MINIMAX_M2,      // Stream a synthetic opening <think> tag and keep </think> tags in `message.content` for MiniMax-M2 compatibility
     // do not extend this enum unless you absolutely have to
     // in most cases, use COMMON_REASONING_FORMAT_AUTO
     // see: https://github.com/ggml-org/llama.cpp/pull/15408

tools/server/README.md

@@ -190,7 +190,7 @@ The project is under active development, and we are [looking for feedback and co
 | `--no-slots` | disables slots monitoring endpoint<br/>(env: LLAMA_ARG_NO_ENDPOINT_SLOTS) |
 | `--slot-save-path PATH` | path to save slot kv cache (default: disabled) |
 | `--jinja` | use jinja template for chat (default: disabled)<br/>(env: LLAMA_ARG_JINJA) |
-| `--reasoning-format FORMAT` | controls whether thought tags are allowed and/or extracted from the response, and in which format they're returned; one of:<br/>- none: leaves thoughts unparsed in `message.content`<br/>- deepseek: puts thoughts in `message.reasoning_content`<br/>- deepseek-legacy: keeps `<think>` tags in `message.content` while also populating `message.reasoning_content`<br/>(default: deepseek)<br/>(env: LLAMA_ARG_THINK) |
+| `--reasoning-format FORMAT` | controls whether thought tags are allowed and/or extracted from the response, and in which format they're returned; one of:<br/>- none: leaves thoughts unparsed in `message.content`<br/>- deepseek: puts thoughts in `message.reasoning_content`<br/>- deepseek-legacy: keeps `<think>` tags in `message.content` while also populating `message.reasoning_content`<br/>- minimax-m2: Stream a synthetic opening <think> tag and keep </think> tags in `message.content` for MiniMax-M2 compatibility<br/>(default: deepseek)<br/>(env: LLAMA_ARG_THINK) |
 | `--reasoning-budget N` | controls the amount of thinking allowed; currently only one of: -1 for unrestricted thinking budget, or 0 to disable thinking (default: -1)<br/>(env: LLAMA_ARG_THINK_BUDGET) |
 | `--chat-template JINJA_TEMPLATE` | set custom jinja chat template (default: template taken from model's metadata)<br/>if suffix/prefix are specified, template will be disabled<br/>only commonly used templates are accepted (unless --jinja is set before this flag):<br/>list of built-in templates:<br/>bailing, chatglm3, chatglm4, chatml, command-r, deepseek, deepseek2, deepseek3, exaone3, exaone4, falcon3, gemma, gigachat, glmedge, gpt-oss, granite, hunyuan-dense, hunyuan-moe, kimi-k2, llama2, llama2-sys, llama2-sys-bos, llama2-sys-strip, llama3, llama4, megrez, minicpm, mistral-v1, mistral-v3, mistral-v3-tekken, mistral-v7, mistral-v7-tekken, monarch, openchat, orion, phi3, phi4, rwkv-world, seed_oss, smolvlm, vicuna, vicuna-orca, yandex, zephyr<br/>(env: LLAMA_ARG_CHAT_TEMPLATE) |
 | `--chat-template-file JINJA_TEMPLATE_FILE` | set custom jinja chat template file (default: template taken from model's metadata)<br/>if suffix/prefix are specified, template will be disabled<br/>only commonly used templates are accepted (unless --jinja is set before this flag):<br/>list of built-in templates:<br/>bailing, chatglm3, chatglm4, chatml, command-r, deepseek, deepseek2, deepseek3, exaone3, exaone4, falcon3, gemma, gigachat, glmedge, gpt-oss, granite, hunyuan-dense, hunyuan-moe, kimi-k2, llama2, llama2-sys, llama2-sys-bos, llama2-sys-strip, llama3, llama4, megrez, minicpm, mistral-v1, mistral-v3, mistral-v3-tekken, mistral-v7, mistral-v7-tekken, monarch, openchat, orion, phi3, phi4, rwkv-world, seed_oss, smolvlm, vicuna, vicuna-orca, yandex, zephyr<br/>(env: LLAMA_ARG_CHAT_TEMPLATE_FILE) |

tools/server/server.cpp

@@ -443,7 +443,10 @@ struct server_task {
             }
             common_reasoning_format reasoning_format = params_base.reasoning_format;
             if (data.contains("reasoning_format")) {
-                reasoning_format = common_reasoning_format_from_name(data.at("reasoning_format").get<std::string>());
+                const auto requested = common_reasoning_format_from_name(data.at("reasoning_format").get<std::string>());
+                if (requested != COMMON_REASONING_FORMAT_AUTO) {
+                    reasoning_format = requested;
+                }
             }
             params.oaicompat_chat_syntax.reasoning_format = reasoning_format;
             params.oaicompat_chat_syntax.reasoning_in_content = params.stream && (reasoning_format == COMMON_REASONING_FORMAT_DEEPSEEK_LEGACY);
@@ -1660,6 +1663,7 @@ struct server_slot {
     bool has_next_token = true;
     bool has_new_line   = false;
     bool truncated      = false;
+    bool minimax_reasoning_prefix_injected = false;
 
     stop_type stop;
 
@@ -1730,6 +1734,7 @@ struct server_slot {
         generated_text = "";
         has_new_line   = false;
         truncated      = false;
+        minimax_reasoning_prefix_injected = false;
         stop           = STOP_TYPE_NONE;
         stopping_word  = "";
         n_sent_text    = 0;
@@ -1856,9 +1861,13 @@ struct server_slot {
         GGML_ASSERT(task);
 
         auto previous_msg = chat_msg;
-        SRV_DBG("Parsing chat message: %s\n", generated_text.c_str());
+        std::string text_to_parse = generated_text;
+        if (minimax_reasoning_prefix_injected) {
+            text_to_parse.insert(0, "<think>\n");
+        }
+        SRV_DBG("Parsing chat message: %s\n", text_to_parse.c_str());
         auto new_msg = common_chat_parse(
-            generated_text,
+            text_to_parse,
             /* is_partial= */ stop != STOP_TYPE_EOS,
             task->params.oaicompat_chat_syntax);
         if (!new_msg.empty()) {
@@ -2793,6 +2802,19 @@ struct server_context {
 
         slot.state = SLOT_STATE_STARTED;
 
+        const bool needs_minimax_prefix =
+            slot.task->params.oaicompat_chat_syntax.reasoning_format == COMMON_REASONING_FORMAT_MINIMAX_M2;
+        if (needs_minimax_prefix) {
+            slot.minimax_reasoning_prefix_injected = true;
+            if (slot.task->params.stream) {
+                completion_token_output prefix_chunk{};
+                prefix_chunk.tok          = LLAMA_TOKEN_NULL;
+                prefix_chunk.prob         = 0.0f;
+                prefix_chunk.text_to_send = "<think>\n";
+                send_partial_response(slot, prefix_chunk, false);
+            }
+        }
+
         SLT_INF(slot, "%s", "processing task\n");
 
         return true;
@@ -2848,7 +2870,10 @@ struct server_context {
                 result.text_to_send = "";
             }
 
+            std::string delta_to_send = result.text_to_send;
+            result.text_to_send = token_str;
             slot.add_token(result);
+            result.text_to_send = std::move(delta_to_send);
             if (slot.task->params.stream) {
                 send_partial_response(slot, result, false);
             }
@@ -3021,7 +3046,11 @@ struct server_context {
         return true;
     }
 
-    void send_partial_response(server_slot & slot, const completion_token_output & tkn, bool is_progress) {
+    void send_partial_response(
+            server_slot & slot,
+            const completion_token_output & tkn,
+            bool is_progress,
+            const std::vector<common_chat_msg_diff> * forced_diffs = nullptr) {
         auto res = std::make_unique<server_task_result_cmpl_partial>();
 
         res->id    = slot.task->id;
@@ -3035,9 +3064,15 @@ struct server_context {
             res->progress.time_ms   = (ggml_time_us() - slot.t_start_process_prompt / 1000);
         } else {
             res->content = tkn.text_to_send;
-            res->tokens  = { tkn.tok };
+            if (tkn.tok != LLAMA_TOKEN_NULL) {
+                res->tokens = { tkn.tok };
+            }
 
-            slot.update_chat_msg(res->oaicompat_msg_diffs);
+            if (forced_diffs) {
+                res->oaicompat_msg_diffs = *forced_diffs;
+            } else {
+                slot.update_chat_msg(res->oaicompat_msg_diffs);
+            }
         }
 
         res->n_decoded           = slot.n_decoded;
@@ -3050,7 +3085,7 @@ struct server_context {
         res->oaicompat_cmpl_id = slot.task->params.oaicompat_cmpl_id;
 
         // populate res.probs_output
-        if (slot.task->params.sampling.n_probs > 0) {
+        if (slot.task->params.sampling.n_probs > 0 && tkn.tok != LLAMA_TOKEN_NULL) {
             res->prob_output = tkn; // copy the token probs
         }
 
@@ -3068,8 +3103,12 @@ struct server_context {
         res->id      = slot.task->id;
         res->id_slot = slot.id;
 
-        res->index           = slot.task->index;
-        res->content         = slot.generated_text;
+        res->index   = slot.task->index;
+        std::string response_content = slot.generated_text;
+        if (slot.minimax_reasoning_prefix_injected) {
+            response_content.insert(0, "<think>\n");
+        }
+        res->content         = std::move(response_content);
         res->tokens          = std::move(slot.generated_tokens);
         res->timings         = slot.get_timings();
         res->prompt          = slot.task->tokens.detokenize(ctx, true);