feat(api): support storing chat completions, enabling evals and model distillation in the dashboard (#72)

Learn more at http://openai.com/devday2024

Author: stainless-app[bot]

.stats.yml

@@ -1,2 +1,2 @@
 configured_endpoints: 68
-openapi_spec_url: https://storage.googleapis.com/stainless-sdk-openapi-specs/openai-8ad878332083dd506a478a293db78dc9e7b1b2124f2682e1d991225bc5bbcc3b.yml
+openapi_spec_url: https://storage.googleapis.com/stainless-sdk-openapi-specs/openai-52b934aee6468039ec7f4ce046a282b5fbce114afc708e70f17121df654f71da.yml

chat.go

@@ -30,33 +30,34 @@ func NewChatService(opts ...option.RequestOption) (r *ChatService) {
 type ChatModel = string
 
 const (
-	ChatModelO1Preview           ChatModel = "o1-preview"
-	ChatModelO1Preview2024_09_12 ChatModel = "o1-preview-2024-09-12"
-	ChatModelO1Mini              ChatModel = "o1-mini"
-	ChatModelO1Mini2024_09_12    ChatModel = "o1-mini-2024-09-12"
-	ChatModelGPT4o               ChatModel = "gpt-4o"
-	ChatModelGPT4o2024_08_06     ChatModel = "gpt-4o-2024-08-06"
-	ChatModelGPT4o2024_05_13     ChatModel = "gpt-4o-2024-05-13"
-	ChatModelChatgpt4oLatest     ChatModel = "chatgpt-4o-latest"
-	ChatModelGPT4oMini           ChatModel = "gpt-4o-mini"
-	ChatModelGPT4oMini2024_07_18 ChatModel = "gpt-4o-mini-2024-07-18"
-	ChatModelGPT4Turbo           ChatModel = "gpt-4-turbo"
-	ChatModelGPT4Turbo2024_04_09 ChatModel = "gpt-4-turbo-2024-04-09"
-	ChatModelGPT4_0125Preview    ChatModel = "gpt-4-0125-preview"
-	ChatModelGPT4TurboPreview    ChatModel = "gpt-4-turbo-preview"
-	ChatModelGPT4_1106Preview    ChatModel = "gpt-4-1106-preview"
-	ChatModelGPT4VisionPreview   ChatModel = "gpt-4-vision-preview"
-	ChatModelGPT4                ChatModel = "gpt-4"
-	ChatModelGPT4_0314           ChatModel = "gpt-4-0314"
-	ChatModelGPT4_0613           ChatModel = "gpt-4-0613"
-	ChatModelGPT4_32k            ChatModel = "gpt-4-32k"
-	ChatModelGPT4_32k0314        ChatModel = "gpt-4-32k-0314"
-	ChatModelGPT4_32k0613        ChatModel = "gpt-4-32k-0613"
-	ChatModelGPT3_5Turbo         ChatModel = "gpt-3.5-turbo"
-	ChatModelGPT3_5Turbo16k      ChatModel = "gpt-3.5-turbo-16k"
-	ChatModelGPT3_5Turbo0301     ChatModel = "gpt-3.5-turbo-0301"
-	ChatModelGPT3_5Turbo0613     ChatModel = "gpt-3.5-turbo-0613"
-	ChatModelGPT3_5Turbo1106     ChatModel = "gpt-3.5-turbo-1106"
-	ChatModelGPT3_5Turbo0125     ChatModel = "gpt-3.5-turbo-0125"
-	ChatModelGPT3_5Turbo16k0613  ChatModel = "gpt-3.5-turbo-16k-0613"
+	ChatModelO1Preview                      ChatModel = "o1-preview"
+	ChatModelO1Preview2024_09_12            ChatModel = "o1-preview-2024-09-12"
+	ChatModelO1Mini                         ChatModel = "o1-mini"
+	ChatModelO1Mini2024_09_12               ChatModel = "o1-mini-2024-09-12"
+	ChatModelGPT4o                          ChatModel = "gpt-4o"
+	ChatModelGPT4o2024_08_06                ChatModel = "gpt-4o-2024-08-06"
+	ChatModelGPT4o2024_05_13                ChatModel = "gpt-4o-2024-05-13"
+	ChatModelGPT4oRealtimePreview2024_10_01 ChatModel = "gpt-4o-realtime-preview-2024-10-01"
+	ChatModelChatgpt4oLatest                ChatModel = "chatgpt-4o-latest"
+	ChatModelGPT4oMini                      ChatModel = "gpt-4o-mini"
+	ChatModelGPT4oMini2024_07_18            ChatModel = "gpt-4o-mini-2024-07-18"
+	ChatModelGPT4Turbo                      ChatModel = "gpt-4-turbo"
+	ChatModelGPT4Turbo2024_04_09            ChatModel = "gpt-4-turbo-2024-04-09"
+	ChatModelGPT4_0125Preview               ChatModel = "gpt-4-0125-preview"
+	ChatModelGPT4TurboPreview               ChatModel = "gpt-4-turbo-preview"
+	ChatModelGPT4_1106Preview               ChatModel = "gpt-4-1106-preview"
+	ChatModelGPT4VisionPreview              ChatModel = "gpt-4-vision-preview"
+	ChatModelGPT4                           ChatModel = "gpt-4"
+	ChatModelGPT4_0314                      ChatModel = "gpt-4-0314"
+	ChatModelGPT4_0613                      ChatModel = "gpt-4-0613"
+	ChatModelGPT4_32k                       ChatModel = "gpt-4-32k"
+	ChatModelGPT4_32k0314                   ChatModel = "gpt-4-32k-0314"
+	ChatModelGPT4_32k0613                   ChatModel = "gpt-4-32k-0613"
+	ChatModelGPT3_5Turbo                    ChatModel = "gpt-3.5-turbo"
+	ChatModelGPT3_5Turbo16k                 ChatModel = "gpt-3.5-turbo-16k"
+	ChatModelGPT3_5Turbo0301                ChatModel = "gpt-3.5-turbo-0301"
+	ChatModelGPT3_5Turbo0613                ChatModel = "gpt-3.5-turbo-0613"
+	ChatModelGPT3_5Turbo1106                ChatModel = "gpt-3.5-turbo-1106"
+	ChatModelGPT3_5Turbo0125                ChatModel = "gpt-3.5-turbo-0125"
+	ChatModelGPT3_5Turbo16k0613             ChatModel = "gpt-3.5-turbo-16k-0613"
 )

chatcompletion.go

@@ -1442,8 +1442,12 @@ func (r ChatCompletionUserMessageParamRole) IsKnown() bool {
 }
 
 type ChatCompletionNewParams struct {
-	// A list of messages comprising the conversation so far.
-	// [Example Python code](https://cookbook.openai.com/examples/how_to_format_inputs_to_chatgpt_models).
+	// A list of messages comprising the conversation so far. Depending on the
+	// [model](https://platform.openai.com/docs/models) you use, different message
+	// types (modalities) are supported, like
+	// [text](https://platform.openai.com/docs/guides/text-generation),
+	// [images](https://platform.openai.com/docs/guides/vision), and
+	// [audio](https://platform.openai.com/docs/guides/audio).
 	Messages param.Field[[]ChatCompletionMessageParamUnion] `json:"messages,required"`
 	// ID of the model to use. See the
 	// [model endpoint compatibility](https://platform.openai.com/docs/models/model-endpoint-compatibility)
@@ -1495,6 +1499,9 @@ type ChatCompletionNewParams struct {
 	// compatible with
 	// [o1 series models](https://platform.openai.com/docs/guides/reasoning).
 	MaxTokens param.Field[int64] `json:"max_tokens"`
+	// Developer-defined tags and values used for filtering completions in the
+	// [dashboard](https://platform.openai.com/completions).
+	Metadata param.Field[map[string]string] `json:"metadata"`
 	// How many chat completion choices to generate for each input message. Note that
 	// you will be charged based on the number of generated tokens across all of the
 	// choices. Keep `n` as `1` to minimize costs.
@@ -1554,6 +1561,9 @@ type ChatCompletionNewParams struct {
 	ServiceTier param.Field[ChatCompletionNewParamsServiceTier] `json:"service_tier"`
 	// Up to 4 sequences where the API will stop generating further tokens.
 	Stop param.Field[ChatCompletionNewParamsStopUnion] `json:"stop"`
+	// Whether or not to store the output of this completion request for traffic
+	// logging in the [dashboard](https://platform.openai.com/completions).
+	Store param.Field[bool] `json:"store"`
 	// Options for streaming response. Only set this when you set `stream: true`.
 	StreamOptions param.Field[ChatCompletionStreamOptionsParam] `json:"stream_options"`
 	// What sampling temperature to use, between 0 and 2. Higher values like 0.8 will

chatcompletion_test.go

@@ -48,15 +48,19 @@ func TestChatCompletionNewWithOptionalParams(t *testing.T) {
 		Logprobs:            openai.F(true),
 		MaxCompletionTokens: openai.F(int64(0)),
 		MaxTokens:           openai.F(int64(0)),
-		N:                   openai.F(int64(1)),
-		ParallelToolCalls:   openai.F(true),
-		PresencePenalty:     openai.F(-2.000000),
+		Metadata: openai.F(map[string]string{
+			"foo": "string",
+		}),
+		N:                 openai.F(int64(1)),
+		ParallelToolCalls: openai.F(true),
+		PresencePenalty:   openai.F(-2.000000),
 		ResponseFormat: openai.F[openai.ChatCompletionNewParamsResponseFormatUnion](shared.ResponseFormatTextParam{
 			Type: openai.F(shared.ResponseFormatTextTypeText),
 		}),
 		Seed:        openai.F(int64(-9007199254740991)),
 		ServiceTier: openai.F(openai.ChatCompletionNewParamsServiceTierAuto),
 		Stop:        openai.F[openai.ChatCompletionNewParamsStopUnion](shared.UnionString("string")),
+		Store:       openai.F(true),
 		StreamOptions: openai.F(openai.ChatCompletionStreamOptionsParam{
 			IncludeUsage: openai.F(true),
 		}),

completion.go

@@ -200,7 +200,9 @@ type CompletionUsage struct {
 	TotalTokens int64 `json:"total_tokens,required"`
 	// Breakdown of tokens used in a completion.
 	CompletionTokensDetails CompletionUsageCompletionTokensDetails `json:"completion_tokens_details"`
-	JSON                    completionUsageJSON                    `json:"-"`
+	// Breakdown of tokens used in the prompt.
+	PromptTokensDetails CompletionUsagePromptTokensDetails `json:"prompt_tokens_details"`
+	JSON                completionUsageJSON                `json:"-"`
 }
 
 // completionUsageJSON contains the JSON metadata for the struct [CompletionUsage]
@@ -209,6 +211,7 @@ type completionUsageJSON struct {
 	PromptTokens            apijson.Field
 	TotalTokens             apijson.Field
 	CompletionTokensDetails apijson.Field
+	PromptTokensDetails     apijson.Field
 	raw                     string
 	ExtraFields             map[string]apijson.Field
 }
@@ -223,6 +226,8 @@ func (r completionUsageJSON) RawJSON() string {
 
 // Breakdown of tokens used in a completion.
 type CompletionUsageCompletionTokensDetails struct {
+	// Audio input tokens generated by the model.
+	AudioTokens int64 `json:"audio_tokens"`
 	// Tokens generated by the model for reasoning.
 	ReasoningTokens int64                                      `json:"reasoning_tokens"`
 	JSON            completionUsageCompletionTokensDetailsJSON `json:"-"`
@@ -231,6 +236,7 @@ type CompletionUsageCompletionTokensDetails struct {
 // completionUsageCompletionTokensDetailsJSON contains the JSON metadata for the
 // struct [CompletionUsageCompletionTokensDetails]
 type completionUsageCompletionTokensDetailsJSON struct {
+	AudioTokens     apijson.Field
 	ReasoningTokens apijson.Field
 	raw             string
 	ExtraFields     map[string]apijson.Field
@@ -244,6 +250,32 @@ func (r completionUsageCompletionTokensDetailsJSON) RawJSON() string {
 	return r.raw
 }
 
+// Breakdown of tokens used in the prompt.
+type CompletionUsagePromptTokensDetails struct {
+	// Audio input tokens present in the prompt.
+	AudioTokens int64 `json:"audio_tokens"`
+	// Cached tokens present in the prompt.
+	CachedTokens int64                                  `json:"cached_tokens"`
+	JSON         completionUsagePromptTokensDetailsJSON `json:"-"`
+}
+
+// completionUsagePromptTokensDetailsJSON contains the JSON metadata for the struct
+// [CompletionUsagePromptTokensDetails]
+type completionUsagePromptTokensDetailsJSON struct {
+	AudioTokens  apijson.Field
+	CachedTokens apijson.Field
+	raw          string
+	ExtraFields  map[string]apijson.Field
+}
+
+func (r *CompletionUsagePromptTokensDetails) UnmarshalJSON(data []byte) (err error) {
+	return apijson.UnmarshalRoot(data, r)
+}
+
+func (r completionUsagePromptTokensDetailsJSON) RawJSON() string {
+	return r.raw
+}
+
 type CompletionNewParams struct {
 	// ID of the model to use. You can use the
 	// [List models](https://platform.openai.com/docs/api-reference/models/list) API to