feat: revamp DeepSeek v4 model handling

- replace legacy DeepSeek ids with the new deepseek-v4 model family\n- move thinking control to request parameters and preserve assistant reasoning content\n- switch history split to IGNORE transcript injection and map upload auth failures to 401\n- update admin defaults, API docs, samples, and tests for the new model scheme

internal/util/standard_request_test.go

@@ -10,9 +10,9 @@ func TestStandardRequestCompletionPayloadSetsModelTypeFromResolvedModel(t *testi
 		search    bool
 		modelType string
 	}{
-		{name: "default", model: "deepseek-chat", thinking: false, search: false, modelType: "default"},
-		{name: "expert", model: "deepseek-expert-reasoner", thinking: true, search: false, modelType: "expert"},
-		{name: "vision", model: "deepseek-vision-chat-search", thinking: false, search: true, modelType: "vision"},
+		{name: "default", model: "deepseek-v4-flash", thinking: false, search: false, modelType: "default"},
+		{name: "expert", model: "deepseek-v4-pro", thinking: true, search: false, modelType: "expert"},
+		{name: "vision", model: "deepseek-v4-vision-search", thinking: false, search: true, modelType: "vision"},
 	}
 
 	for _, tc := range tests {

internal/util/thinking.go

@@ -0,0 +1,53 @@
+package util
+
+import "strings"
+
+func ResolveThinkingEnabled(req map[string]any, defaultEnabled bool) bool {
+	if enabled, ok := parseThinkingSetting(req["thinking"]); ok {
+		return enabled
+	}
+	if extraBody, ok := req["extra_body"].(map[string]any); ok {
+		if enabled, ok := parseThinkingSetting(extraBody["thinking"]); ok {
+			return enabled
+		}
+	}
+	if enabled, ok := parseReasoningEffort(req["reasoning_effort"]); ok {
+		return enabled
+	}
+	return defaultEnabled
+}
+
+func parseThinkingSetting(raw any) (bool, bool) {
+	switch v := raw.(type) {
+	case string:
+		switch strings.ToLower(strings.TrimSpace(v)) {
+		case "enabled":
+			return true, true
+		case "disabled":
+			return false, true
+		default:
+			return false, false
+		}
+	case map[string]any:
+		if typ, ok := v["type"]; ok {
+			return parseThinkingSetting(typ)
+		}
+	}
+	return false, false
+}
+
+func parseReasoningEffort(raw any) (bool, bool) {
+	switch strings.ToLower(strings.TrimSpace(toString(raw))) {
+	case "low", "medium", "high", "xhigh":
+		return true, true
+	default:
+		return false, false
+	}
+}
+
+func toString(raw any) string {
+	if s, ok := raw.(string); ok {
+		return s
+	}
+	return ""
+}

internal/util/thinking_test.go

@@ -0,0 +1,44 @@
+package util
+
+import "testing"
+
+func TestResolveThinkingEnabledPriority(t *testing.T) {
+	req := map[string]any{
+		"thinking": map[string]any{"type": "disabled"},
+		"extra_body": map[string]any{
+			"thinking": map[string]any{"type": "enabled"},
+		},
+		"reasoning_effort": "high",
+	}
+	if got := ResolveThinkingEnabled(req, true); got {
+		t.Fatalf("expected top-level thinking to win, got enabled=%v", got)
+	}
+}
+
+func TestResolveThinkingEnabledUsesExtraBodyFallback(t *testing.T) {
+	req := map[string]any{
+		"extra_body": map[string]any{
+			"thinking": map[string]any{"type": "disabled"},
+		},
+	}
+	if got := ResolveThinkingEnabled(req, true); got {
+		t.Fatalf("expected extra_body thinking to disable, got enabled=%v", got)
+	}
+}
+
+func TestResolveThinkingEnabledMapsReasoningEffortToEnabled(t *testing.T) {
+	for _, effort := range []string{"low", "medium", "high", "xhigh"} {
+		if got := ResolveThinkingEnabled(map[string]any{"reasoning_effort": effort}, false); !got {
+			t.Fatalf("expected reasoning_effort=%s to enable thinking", effort)
+		}
+	}
+}
+
+func TestResolveThinkingEnabledDefaultsWhenUnset(t *testing.T) {
+	if !ResolveThinkingEnabled(nil, true) {
+		t.Fatal("expected default thinking=true when unset")
+	}
+	if ResolveThinkingEnabled(nil, false) {
+		t.Fatal("expected default thinking=false when unset")
+	}
+}

internal/util/util_edge_test.go

@@ -349,14 +349,14 @@ func TestConvertClaudeToDeepSeekNoSystem(t *testing.T) {
 }
 
 func TestConvertClaudeToDeepSeekOpusUsesSlowMapping(t *testing.T) {
-	t.Setenv("DS2API_CONFIG_JSON", `{"keys":[],"accounts":[],"claude_mapping":{"fast":"deepseek-chat","slow":"deepseek-reasoner"}}`)
+	t.Setenv("DS2API_CONFIG_JSON", `{"keys":[],"accounts":[],"claude_mapping":{"fast":"deepseek-v4-flash","slow":"deepseek-v4-pro"}}`)
 	store := config.LoadStore()
 	req := map[string]any{
 		"model":    "claude-opus-4-6",
 		"messages": []any{map[string]any{"role": "user", "content": "Hi"}},
 	}
 	out := ConvertClaudeToDeepSeek(req, store)
-	if out["model"] != "deepseek-reasoner" {
+	if out["model"] != "deepseek-v4-pro" {
 		t.Fatalf("expected opus to use slow mapping, got %q", out["model"])
 	}
 }