add properties to translation & assistant API

Author: michal.zyga

core/src/main/scala/sttp/openai/OpenAI.scala

@@ -463,7 +463,8 @@ class OpenAI(authToken: String, baseUri: Uri = OpenAIUris.OpenAIBaseUri) {
           Some(multipart("model", model.value)),
           prompt.map(multipart("prompt", _)),
           responseFormat.map(format => multipart("response_format", format)),
-          temperature.map(multipart("temperature", _))
+          temperature.map(multipart("temperature", _)),
+          language.map(multipart("language", _))
         ).flatten
       }
       .response(asJson_parseErrors[AudioResponse])

core/src/main/scala/sttp/openai/requests/assistants/AssistantsRequestBody.scala

@@ -8,27 +8,31 @@ object AssistantsRequestBody {
   /** @param model
     *   ID of the model to use. You can use the List models API to see all of your available models, or see our Model overview for
     *   descriptions of them.
-    *
     * @param name
     *   The name of the assistant. The maximum length is 256 characters.
-    *
     * @param description
     *   The description of the assistant. The maximum length is 512 characters.
-    *
     * @param instructions
     *   The system instructions that the assistant uses. The maximum length is 32768 characters.
-    *
+    * @param reasoningEffort
+    *   o1 and o3-mini models only Constrains effort on reasoning for reasoning models. Currently supported values are low, medium, and
+    *   high. Reducing reasoning effort can result in faster responses and fewer tokens used on reasoning in a response.
     * @param tools
     *   A list of tool enabled on the assistant. There can be a maximum of 128 tools per assistant. Tools can be of types code_interpreter,
     *   file_search, or function.
-    *
     * @param toolResources
     *   A set of resources that are used by the assistant's tools. The resources are specific to the type of tool. For example, the
     *   code_interpreter tool requires a list of file IDs, while the file_search tool requires a list of vector store IDs.
-    *
     * @param metadata
     *   Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object
     *   in a structured format. Keys can be a maximum of 64 characters long and values can be a maxium of 512 characters long.
+    * @param temperature
+    *   What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like
+    *   0.2 will make it more focused and deterministic.
+    * @param topP
+    *   An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p
+    *   probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered. We generally recommend
+    *   altering this or temperature but not both.
     *
     * For more information please visit: [[https://platform.openai.com/docs/api-reference/assistants/createAssistant]]
     */
@@ -37,9 +41,12 @@ object AssistantsRequestBody {
       name: Option[String] = None,
       description: Option[String] = None,
       instructions: Option[String] = None,
+      reasoningEffort: Option[ReasoningEffort] = None,
       tools: Seq[Tool] = Seq.empty,
       toolResources: Option[ToolResources] = None,
-      metadata: Option[Map[String, String]] = None
+      metadata: Option[Map[String, String]] = None,
+      temperature: Option[Float] = None,
+      topP: Option[Float] = None
   )
   object CreateAssistantBody {
     implicit val createAssistantBodyW: SnakePickle.Writer[CreateAssistantBody] = SnakePickle.macroW[CreateAssistantBody]
@@ -48,26 +55,33 @@ object AssistantsRequestBody {
   /** @param model
     *   ID of the model to use. You can use the List models API to see all of your available models, or see our Model overview for
     *   descriptions of them.
-    *
     * @param name
     *   The name of the assistant. The maximum length is 256 characters.
-    *
     * @param description
     *   The description of the assistant. The maximum length is 512 characters.
-    *
     * @param instructions
     *   The system instructions that the assistant uses. The maximum length is 32768 characters.
+    * @param reasoningEffort
+    *   o1 and o3-mini models only
     *
+    * Constrains effort on reasoning for reasoning models. Currently supported values are low, medium, and high. Reducing reasoning effort
+    * can result in faster responses and fewer tokens used on reasoning in a response.
     * @param tools
     *   A list of tool enabled on the assistant. There can be a maximum of 128 tools per assistant. Tools can be of types code_interpreter,
     *   file_search, or function.
-    *
     * @param toolResources
     *   A set of resources that are used by the assistant's tools. The resources are specific to the type of tool. For example, the
     *   code_interpreter tool requires a list of file IDs, while the file_search tool requires a list of vector store IDs. v
     * @param metadata
     *   Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object
     *   in a structured format. Keys can be a maximum of 64 characters long and values can be a maxium of 512 characters long.
+    * @param temperature
+    *   What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like
+    *   0.2 will make it more focused and deterministic.
+    * @param topP
+    *   An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p
+    *   probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered. We generally recommend
+    *   altering this or temperature but not both.
     *
     * For more information please visit: [[https://platform.openai.com/docs/api-reference/assistants/modifyAssistant]]
     */
@@ -76,12 +90,33 @@ object AssistantsRequestBody {
       name: Option[String] = None,
       description: Option[String] = None,
       instructions: Option[String] = None,
+      reasoningEffort: Option[ReasoningEffort] = None,
       tools: Seq[Tool] = Seq.empty,
       toolResources: Option[ToolResources] = None,
-      metadata: Map[String, String] = Map.empty
+      metadata: Map[String, String] = Map.empty,
+      temperature: Option[Float] = None,
+      topP: Option[Float] = None
   )
 
   object ModifyAssistantBody {
     implicit val modifyAssistantBodyW: SnakePickle.Writer[ModifyAssistantBody] = SnakePickle.macroW[ModifyAssistantBody]
   }
 }
+
+sealed abstract class ReasoningEffort(val value: String)
+
+object ReasoningEffort {
+
+  implicit val reasoningEffortW: SnakePickle.Writer[ReasoningEffort] = SnakePickle
+    .writer[ujson.Value]
+    .comap[ReasoningEffort](reasoningEffort => SnakePickle.writeJs(reasoningEffort.value))
+
+  case object Low extends ReasoningEffort("low")
+
+  case object Medium extends ReasoningEffort("medium")
+
+  case object High extends ReasoningEffort("high")
+
+  case class CustomReasoningEffort(customReasoningEffort: String) extends ReasoningEffort(customReasoningEffort)
+
+}

core/src/main/scala/sttp/openai/requests/audio/translations/TranslationConfig.scala

@@ -4,7 +4,6 @@ import sttp.openai.requests.audio.RecognitionModel
 import sttp.openai.requests.images.ResponseFormat
 
 import java.io.File
-import java.nio.file.Paths
 
 /** @param file
   *   The audio file to translate, in one of these formats: mp3, mp4, mpeg, mpga, m4a, wav, or webm.
@@ -18,28 +17,14 @@ import java.nio.file.Paths
   *   The sampling temperature, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will
   *   make it more focused and deterministic. If set to 0, the model will use log probability to automatically increase the temperature
   *   until certain thresholds are hit.
+  * @param language
+  *   The language of the input audio. Supplying the input language in ISO-639-1 (e.g. en) format will improve accuracy and latency.
   */
 case class TranslationConfig(
     file: File,
     model: RecognitionModel,
     prompt: Option[String] = None,
     responseFormat: Option[ResponseFormat] = None,
-    temperature: Option[Float] = None
+    temperature: Option[Float] = None,
+    language: Option[String] = None
 )
-
-object TranslationConfig {
-  def createTranslationConfigWithSystemPaths(
-      systemPath: String,
-      model: RecognitionModel,
-      prompt: Option[String] = None,
-      responseFormat: Option[ResponseFormat] = None,
-      temperature: Option[Float] = None
-  ): TranslationConfig =
-    TranslationConfig(
-      Paths.get(systemPath).toFile,
-      model,
-      prompt,
-      responseFormat,
-      temperature
-    )
-}

core/src/main/scala/sttp/openai/requests/embeddings/EmbeddingsRequestBody.scala

@@ -12,8 +12,10 @@ object EmbeddingsRequestBody {
     *   Input text to get embeddings for, encoded as a string or array of tokens.
     * @param user
     *   A unique identifier representing your end-user, which can help OpenAI to monitor and detect abuse.
+    * @param dimensions
+    *   The number of dimensions for the embeddings. Only supported in text-embedding-3 and later models.
     */
-  case class EmbeddingsBody(model: EmbeddingsModel, input: EmbeddingsInput, user: Option[String] = None)
+  case class EmbeddingsBody(model: EmbeddingsModel, input: EmbeddingsInput, user: Option[String] = None, dimensions: Option[Int] = None)
 
   object EmbeddingsBody {
     implicit val embeddingsBodyWriter: SnakePickle.Writer[EmbeddingsBody] = SnakePickle.macroW

core/src/test/scala/sttp/openai/fixtures/AssistantsFixture.scala

@@ -7,7 +7,10 @@ object AssistantsFixture {
       |  "instructions": "You are a personal math tutor. When asked a question, write and run Python code to answer the question.",
       |  "name": "Math Tutor",
       |  "tools": [{"type": "code_interpreter"}],
-      |  "model": "gpt-4"
+      |  "model": "gpt-4",
+      |  "reasoning_effort": "low",
+      |  "temperature": 1.0,
+      |  "top_p": 1.0
       |}""".stripMargin
 
   val jsonCreateAssistantResponse: String =
@@ -113,7 +116,10 @@ object AssistantsFixture {
       |     "file_search": {
       |        "vector_store_ids": ["vs_1", "vs_3"]
       |    }
-      |  }
+      |  },
+      |  "reasoning_effort": "low",
+      |  "temperature": 1.0,
+      |  "top_p": 1.0
       |}
       |""".stripMargin
 

core/src/test/scala/sttp/openai/requests/assistants/AssistantsDataSpec.scala

@@ -20,7 +20,10 @@ class AssistantsDataSpec extends AnyFlatSpec with Matchers with EitherValues {
       instructions = Some("You are a personal math tutor. When asked a question, write and run Python code to answer the question."),
       name = Some("Math Tutor"),
       tools = Seq(CodeInterpreterTool),
-      model = "gpt-4"
+      model = "gpt-4",
+      reasoningEffort = Some(ReasoningEffort.Low),
+      temperature = Some(1.0f),
+      topP = Some(1.0f)
     )
 
     val jsonRequest: ujson.Value = ujson.read(fixtures.AssistantsFixture.jsonCreateAssistantRequest)
@@ -213,7 +216,10 @@ class AssistantsDataSpec extends AnyFlatSpec with Matchers with EitherValues {
       ),
       tools = Seq(FileSearchTool),
       model = Some("gpt-4"),
-      toolResources = Some(ToolResources(None, Some(FileSearchToolResource(Some(Seq("vs_1", "vs_3"))))))
+      toolResources = Some(ToolResources(None, Some(FileSearchToolResource(Some(Seq("vs_1", "vs_3")))))),
+      reasoningEffort = Some(ReasoningEffort.Low),
+      temperature = Some(1.0f),
+      topP = Some(1.0f)
     )
 
     val jsonRequest: ujson.Value = ujson.read(fixtures.AssistantsFixture.jsonModifyAssistantRequest)