feat: support openai websearch models

relay/model/general.go

@@ -25,46 +25,50 @@ type StreamOptions struct {
 
 type GeneralOpenAIRequest struct {
 	// https://platform.openai.com/docs/api-reference/chat/create
-	Messages            []Message       `json:"messages,omitempty"`
-	Model               string          `json:"model,omitempty"`
-	Store               *bool           `json:"store,omitempty"`
-	ReasoningEffort     *string         `json:"reasoning_effort,omitempty"`
-	Metadata            any             `json:"metadata,omitempty"`
-	FrequencyPenalty    *float64        `json:"frequency_penalty,omitempty"`
-	LogitBias           any             `json:"logit_bias,omitempty"`
-	Logprobs            *bool           `json:"logprobs,omitempty"`
-	TopLogprobs         *int            `json:"top_logprobs,omitempty"`
-	MaxTokens           int             `json:"max_tokens,omitempty"`
-	MaxCompletionTokens *int            `json:"max_completion_tokens,omitempty"`
-	N                   int             `json:"n,omitempty"`
-	Modalities          []string        `json:"modalities,omitempty"`
-	Prediction          any             `json:"prediction,omitempty"`
-	Audio               *Audio          `json:"audio,omitempty"`
-	PresencePenalty     *float64        `json:"presence_penalty,omitempty"`
-	ResponseFormat      *ResponseFormat `json:"response_format,omitempty"`
-	Seed                float64         `json:"seed,omitempty"`
-	ServiceTier         *string         `json:"service_tier,omitempty"`
-	Stop                any             `json:"stop,omitempty"`
-	Stream              bool            `json:"stream,omitempty"`
-	StreamOptions       *StreamOptions  `json:"stream_options,omitempty"`
-	Temperature         *float64        `json:"temperature,omitempty"`
-	TopP                *float64        `json:"top_p,omitempty"`
-	TopK                int             `json:"top_k,omitempty"`
-	Tools               []Tool          `json:"tools,omitempty"`
-	ToolChoice          any             `json:"tool_choice,omitempty"`
-	ParallelTooCalls    *bool           `json:"parallel_tool_calls,omitempty"`
-	User                string          `json:"user,omitempty"`
-	FunctionCall        any             `json:"function_call,omitempty"`
-	Functions           any             `json:"functions,omitempty"`
+	Messages            []Message `json:"messages,omitempty"`
+	Model               string    `json:"model,omitempty"`
+	Store               *bool     `json:"store,omitempty"`
+	Metadata            any       `json:"metadata,omitempty"`
+	FrequencyPenalty    *float64  `json:"frequency_penalty,omitempty"`
+	LogitBias           any       `json:"logit_bias,omitempty"`
+	Logprobs            *bool     `json:"logprobs,omitempty"`
+	TopLogprobs         *int      `json:"top_logprobs,omitempty"`
+	MaxTokens           int       `json:"max_tokens,omitempty"`
+	MaxCompletionTokens *int      `json:"max_completion_tokens,omitempty"`
+	N                   int       `json:"n,omitempty"`
+	// ReasoningEffort constrains effort on reasoning for reasoning models, reasoning models only.
+	ReasoningEffort *string `json:"reasoning_effort,omitempty" binding:"omitempty,oneof=low medium high"`
+	// Modalities currently the model only programmatically allows modalities = [“text”, “audio”]
+	Modalities       []string        `json:"modalities,omitempty"`
+	Prediction       any             `json:"prediction,omitempty"`
+	Audio            *Audio          `json:"audio,omitempty"`
+	PresencePenalty  *float64        `json:"presence_penalty,omitempty"`
+	ResponseFormat   *ResponseFormat `json:"response_format,omitempty"`
+	Seed             float64         `json:"seed,omitempty"`
+	ServiceTier      *string         `json:"service_tier,omitempty" binding:"omitempty,oneof=default auto"`
+	Stop             any             `json:"stop,omitempty"`
+	Stream           bool            `json:"stream,omitempty"`
+	StreamOptions    *StreamOptions  `json:"stream_options,omitempty"`
+	Temperature      *float64        `json:"temperature,omitempty"`
+	TopP             *float64        `json:"top_p,omitempty"`
+	TopK             int             `json:"top_k,omitempty"`
+	Tools            []Tool          `json:"tools,omitempty"`
+	ToolChoice       any             `json:"tool_choice,omitempty"`
+	ParallelTooCalls *bool           `json:"parallel_tool_calls,omitempty"`
+	User             string          `json:"user,omitempty"`
+	FunctionCall     any             `json:"function_call,omitempty"`
+	Functions        any             `json:"functions,omitempty"`
 	// https://platform.openai.com/docs/api-reference/embeddings/create
 	Input          any    `json:"input,omitempty"`
 	EncodingFormat string `json:"encoding_format,omitempty"`
 	Dimensions     int    `json:"dimensions,omitempty"`
 	// https://platform.openai.com/docs/api-reference/images/create
-	Prompt  any     `json:"prompt,omitempty"`
-	Quality *string `json:"quality,omitempty"`
-	Size    string  `json:"size,omitempty"`
-	Style   *string `json:"style,omitempty"`
+	Prompt           string            `json:"prompt,omitempty"`
+	Quality          *string           `json:"quality,omitempty"`
+	Size             string            `json:"size,omitempty"`
+	Style            *string           `json:"style,omitempty"`
+	WebSearchOptions *WebSearchOptions `json:"web_search_options,omitempty"`
+
 	// Others
 	Instruction string `json:"instruction,omitempty"`
 	NumCtx      int    `json:"num_ctx,omitempty"`
@@ -79,6 +83,34 @@ type GeneralOpenAIRequest struct {
 	Thinking *Thinking `json:"thinking,omitempty"`
 }
 
+// WebSearchOptions is the tool searches the web for relevant results to use in a response.
+type WebSearchOptions struct {
+	// SearchContextSize is the high level guidance for the amount of context window space to use for the search,
+	// default is "medium".
+	SearchContextSize *string       `json:"search_context_size,omitempty" binding:"omitempty,oneof=low medium high"`
+	UserLocation      *UserLocation `json:"user_location,omitempty"`
+}
+
+// UserLocation is a struct that contains the location of the user.
+type UserLocation struct {
+	// Approximate is the approximate location parameters for the search.
+	Approximate UserLocationApproximate `json:"approximate" binding:"required"`
+	// Type is the type of location approximation.
+	Type string `json:"type" binding:"required,oneof=approximate"`
+}
+
+// UserLocationApproximate is a struct that contains the approximate location of the user.
+type UserLocationApproximate struct {
+	// City is the city of the user, e.g. San Francisco.
+	City *string `json:"city,omitempty"`
+	// Country is the country of the user, e.g. US.
+	Country *string `json:"country,omitempty"`
+	// Region is the region of the user, e.g. California.
+	Region *string `json:"region,omitempty"`
+	// Timezone is the IANA timezone of the user, e.g. America/Los_Angeles.
+	Timezone *string `json:"timezone,omitempty"`
+}
+
 // https://docs.anthropic.com/en/docs/build-with-claude/extended-thinking#implementing-extended-thinking
 type Thinking struct {
 	Type         string `json:"type"`

relay/model/message.go

@@ -1,13 +1,43 @@
 package model
 
+import (
+	"context"
+	"strings"
+
+	"github.com/songquanpeng/one-api/common/logger"
+)
+
+// ReasoningFormat is the format of reasoning content,
+// can be set by the reasoning_format parameter in the request url.
+type ReasoningFormat string
+
+const (
+	ReasoningFormatUnspecified ReasoningFormat = ""
+	// ReasoningFormatReasoningContent is the reasoning format used by deepseek official API
+	ReasoningFormatReasoningContent ReasoningFormat = "reasoning_content"
+	// ReasoningFormatReasoning is the reasoning format used by openrouter
+	ReasoningFormatReasoning ReasoningFormat = "reasoning"
+
+	// ReasoningFormatThinkTag is the reasoning format used by 3rd party deepseek-r1 providers.
+	//
+	// Deprecated: I believe <think> is a very poor format, especially in stream mode, it is difficult to extract and convert.
+	// Considering that only a few deepseek-r1 third-party providers use this format, it has been decided to no longer support it.
+	// ReasoningFormatThinkTag ReasoningFormat = "think-tag"
+
+	// ReasoningFormatThinking is the reasoning format used by anthropic
+	ReasoningFormatThinking ReasoningFormat = "thinking"
+)
+
 type Message struct {
 	Role string `json:"role,omitempty"`
 	// Content is a string or a list of objects
-	Content    any           `json:"content,omitempty"`
-	Name       *string       `json:"name,omitempty"`
-	ToolCalls  []Tool        `json:"tool_calls,omitempty"`
-	ToolCallId string        `json:"tool_call_id,omitempty"`
-	Audio      *messageAudio `json:"audio,omitempty"`
+	Content    any              `json:"content,omitempty"`
+	Name       *string          `json:"name,omitempty"`
+	ToolCalls  []Tool           `json:"tool_calls,omitempty"`
+	ToolCallId string           `json:"tool_call_id,omitempty"`
+	Audio      *messageAudio    `json:"audio,omitempty"`
+	Annotation []AnnotationItem `json:"annotation,omitempty"`
+
 	// -------------------------------------
 	// Deepseek 专有的一些字段
 	// https://api-docs.deepseek.com/api/create-chat-completion
@@ -18,11 +48,52 @@ type Message struct {
 	// Prefix Completion feature as the input for the CoT in the last assistant message.
 	// When using this feature, the prefix parameter must be set to true.
 	ReasoningContent *string `json:"reasoning_content,omitempty"`
+
 	// -------------------------------------
 	// Openrouter
 	// -------------------------------------
 	Reasoning *string `json:"reasoning,omitempty"`
 	Refusal   *bool   `json:"refusal,omitempty"`
+
+	// -------------------------------------
+	// Anthropic
+	// -------------------------------------
+	Thinking  *string `json:"thinking,omitempty"`
+	Signature *string `json:"signature,omitempty"`
+}
+
+type AnnotationItem struct {
+	Type        string      `json:"type" binding:"oneof=url_citation"`
+	UrlCitation UrlCitation `json:"url_citation"`
+}
+
+// UrlCitation is a URL citation when using web search.
+type UrlCitation struct {
+	// Endpoint is the index of the last character of the URL citation in the message.
+	EndIndex int `json:"end_index"`
+	// StartIndex is the index of the first character of the URL citation in the message.
+	StartIndex int `json:"start_index"`
+	// Title is the title of the web resource.
+	Title string `json:"title"`
+	// Url is the URL of the web resource.
+	Url string `json:"url"`
+}
+
+// SetReasoningContent sets the reasoning content based on the format
+func (m *Message) SetReasoningContent(format string, reasoningContent string) {
+	switch ReasoningFormat(strings.ToLower(strings.TrimSpace(format))) {
+	case ReasoningFormatReasoningContent:
+		m.ReasoningContent = &reasoningContent
+		// case ReasoningFormatThinkTag:
+		// 	m.Content = fmt.Sprintf("<think>%s</think>%s", reasoningContent, m.Content)
+	case ReasoningFormatThinking:
+		m.Thinking = &reasoningContent
+	case ReasoningFormatReasoning,
+		ReasoningFormatUnspecified:
+		m.Reasoning = &reasoningContent
+	default:
+		logger.Warnf(context.TODO(), "unknown reasoning format: %q", format)
+	}
 }
 
 type messageAudio struct {
@@ -50,6 +121,7 @@ func (m Message) StringContent() string {
 			if !ok {
 				continue
 			}
+
 			if contentMap["type"] == ContentTypeText {
 				if subStr, ok := contentMap["text"].(string); ok {
 					contentStr += subStr
@@ -58,6 +130,7 @@ func (m Message) StringContent() string {
 		}
 		return contentStr
 	}
+
 	return ""
 }
 
@@ -71,6 +144,7 @@ func (m Message) ParseContent() []MessageContent {
 		})
 		return contentList
 	}
+
 	anyList, ok := m.Content.([]any)
 	if ok {
 		for _, contentItem := range anyList {
@@ -95,8 +169,21 @@ func (m Message) ParseContent() []MessageContent {
 						},
 					})
 				}
+			case ContentTypeInputAudio:
+				if subObj, ok := contentMap["input_audio"].(map[string]any); ok {
+					contentList = append(contentList, MessageContent{
+						Type: ContentTypeInputAudio,
+						InputAudio: &InputAudio{
+							Data:   subObj["data"].(string),
+							Format: subObj["format"].(string),
+						},
+					})
+				}
+			default:
+				logger.Warnf(context.TODO(), "unknown content type: %s", contentMap["type"])
 			}
 		}
+
 		return contentList
 	}
 	return nil
@@ -108,7 +195,23 @@ type ImageURL struct {
 }
 
 type MessageContent struct {
-	Type     string    `json:"type,omitempty"`
-	Text     string    `json:"text"`
-	ImageURL *ImageURL `json:"image_url,omitempty"`
+	// Type should be one of the following: text/input_audio
+	Type       string      `json:"type,omitempty"`
+	Text       string      `json:"text"`
+	ImageURL   *ImageURL   `json:"image_url,omitempty"`
+	InputAudio *InputAudio `json:"input_audio,omitempty"`
+	// -------------------------------------
+	// Anthropic
+	// -------------------------------------
+	Thinking  *string `json:"thinking,omitempty"`
+	Signature *string `json:"signature,omitempty"`
+}
+
+type InputAudio struct {
+	// Data is the base64 encoded audio data
+	Data string `json:"data" binding:"required"`
+	// Format is the audio format, should be one of the
+	// following: mp3/mp4/mpeg/mpga/m4a/wav/webm/pcm16.
+	// When stream=true, format should be pcm16
+	Format string `json:"format"`
 }

relay/model/misc.go

@@ -1,15 +1,22 @@
 package model
 
+// Usage is the token usage information returned by OpenAI API.
 type Usage struct {
 	PromptTokens     int `json:"prompt_tokens"`
 	CompletionTokens int `json:"completion_tokens"`
 	TotalTokens      int `json:"total_tokens"`
 	// PromptTokensDetails may be empty for some models
-	PromptTokensDetails *usagePromptTokensDetails `gorm:"-" json:"prompt_tokens_details,omitempty"`
+	PromptTokensDetails *usagePromptTokensDetails `json:"prompt_tokens_details,omitempty"`
 	// CompletionTokensDetails may be empty for some models
-	CompletionTokensDetails *usageCompletionTokensDetails `gorm:"-" json:"completion_tokens_details,omitempty"`
-	ServiceTier             string                        `gorm:"-" json:"service_tier,omitempty"`
-	SystemFingerprint       string                        `gorm:"-" json:"system_fingerprint,omitempty"`
+	CompletionTokensDetails *usageCompletionTokensDetails `json:"completion_tokens_details,omitempty"`
+	ServiceTier             string                        `json:"service_tier,omitempty"`
+	SystemFingerprint       string                        `json:"system_fingerprint,omitempty"`
+
+	// -------------------------------------
+	// Custom fields
+	// -------------------------------------
+	// ToolsCost is the cost of using tools, in quota.
+	ToolsCost int64 `json:"tools_cost,omitempty"`
 }
 
 type Error struct {