com.simiacryptus.jopenai.models.ChatModels.kt Maven / Gradle / Ivy
package com.simiacryptus.jopenai.models
import com.fasterxml.jackson.core.JsonGenerator
import com.fasterxml.jackson.core.JsonParser
import com.fasterxml.jackson.databind.DeserializationContext
import com.fasterxml.jackson.databind.JsonDeserializer
import com.fasterxml.jackson.databind.SerializerProvider
import com.fasterxml.jackson.databind.annotation.JsonDeserialize
import com.fasterxml.jackson.databind.annotation.JsonSerialize
import com.fasterxml.jackson.databind.ser.std.StdSerializer
import com.simiacryptus.jopenai.ApiModel.Usage
import com.simiacryptus.jopenai.models.ChatModels.Companion.values
@JsonDeserialize(using = ChatModelsDeserializer::class)
@JsonSerialize(using = ChatModelsSerializer::class)
open class ChatModels(
val name : String,
modelName: String,
maxTokens: Int,
val provider: APIProvider,
private val inputTokenPricePerK: Double,
private val outputTokenPricePerK: Double,
) : OpenAITextModel(modelName, maxTokens) {
override fun pricing(usage: Usage) =
(usage.prompt_tokens * inputTokenPricePerK + usage.completion_tokens * outputTokenPricePerK) / 1000.0
companion object {
val GPT35Turbo = ChatModels(
name = "GPT35Turbo",
modelName = "gpt-3.5-turbo-0125",
maxTokens = 16384,
provider = APIProvider.OpenAI,
inputTokenPricePerK = 0.0005,
outputTokenPricePerK = 0.0015
)
/*
val GPT4 = ChatModels(
name = "GPT4",
modelName = "gpt-4-32k",
maxTokens = 32768,
provider = APIProvider.OpenAI,
inputTokenPricePerK = 0.06,
outputTokenPricePerK = 0.12
)
*/
val GPT4Turbo = ChatModels(
name = "GPT4Turbo",
modelName = "gpt-4-turbo-preview",
maxTokens = 128000,
provider = APIProvider.OpenAI,
inputTokenPricePerK = 0.01,
outputTokenPricePerK = 0.03
)
val GPT4Vision = ChatModels(
name = "GPT4Vision",
modelName = "gpt-4-vision-preview",
maxTokens = 8192,
provider = APIProvider.OpenAI,
inputTokenPricePerK = 0.01,
outputTokenPricePerK = 0.03
)
private val SonarSmallChat = ChatModels(
name = "SonarSmallChat",
modelName = "sonar-small-chat",
maxTokens = 16384,
provider = APIProvider.Perplexity,
inputTokenPricePerK = 0.0005,
outputTokenPricePerK = 0.0015
)
private val SonarSmallOnline = ChatModels(
name = "SonarSmallOnline",
modelName = "sonar-small-online",
maxTokens = 12000,
provider = APIProvider.Perplexity,
inputTokenPricePerK = 0.0005,
outputTokenPricePerK = 0.0015
)
private val SonarMediumChat = ChatModels(
name = "SonarMediumChat",
modelName = "sonar-medium-chat",
maxTokens = 16384,
provider = APIProvider.Perplexity,
inputTokenPricePerK = 0.0005,
outputTokenPricePerK = 0.0015
)
private val SonarMediumOnline = ChatModels(
name = "SonarMediumOnline",
modelName = "sonar-medium-online",
maxTokens = 12000,
provider = APIProvider.Perplexity,
inputTokenPricePerK = 0.0005,
outputTokenPricePerK = 0.0015
)
private val Codellama70bInstruct = ChatModels(
name = "Codellama70bInstruct",
modelName = "codellama-70b-instruct",
maxTokens = 16384,
provider = APIProvider.Perplexity,
inputTokenPricePerK = 0.0005,
outputTokenPricePerK = 0.0015
)
private val Mistral7bInstruct = ChatModels(
name = "Mistral7bInstruct",
modelName = "mistral-7b-instruct",
maxTokens = 16384,
provider = APIProvider.Perplexity,
inputTokenPricePerK = 0.0005,
outputTokenPricePerK = 0.0015
)
private val Mixtral8x7bInstruct = ChatModels(
name = "Mixtral8x7bInstruct",
modelName = "mixtral-8x7b-instruct",
maxTokens = 16384,
provider = APIProvider.Perplexity,
inputTokenPricePerK = 0.0005,
outputTokenPricePerK = 0.0015
)
private val LLaMA270bChat = ChatModels(
name = "LLaMA270bChat",
modelName = "llama2-70b-4096",
maxTokens = 4096,
provider = APIProvider.Groq,
inputTokenPricePerK = 0.0005,
outputTokenPricePerK = 0.0015
)
private val Mixtral8x7bInstructV01 = ChatModels(
name = "Mixtral8x7bInstructV01",
modelName = "mixtral-8x7b-32768",
maxTokens = 32768,
provider = APIProvider.Groq,
inputTokenPricePerK = 0.0005,
outputTokenPricePerK = 0.0015
)
/*
val Gemma7bIt = ChatModels(
name = "Gemma7bIt",
modelName = "Gemma-7b-it",
maxTokens = 8192,
provider = APIProvider.Groq,
inputTokenPricePerK = 0.0005,
outputTokenPricePerK = 0.0015
)
*/
private val Zephyr7bBeta = ChatModels(
name = "Zephyr7bBeta",
modelName = "zephyr-7b-beta",
maxTokens = 16384,
provider = APIProvider.ModelsLab,
inputTokenPricePerK = 0.0005,
outputTokenPricePerK = 0.0015
)
private val DialoGPTLarge = ChatModels(
name = "DialoGPTLarge",
modelName = "DialoGPT-large",
maxTokens = 16384,
provider = APIProvider.ModelsLab,
inputTokenPricePerK = 0.0005,
outputTokenPricePerK = 0.0015
)
private val YarnMistral7b128k = ChatModels(
name = "YarnMistral7b128k",
modelName = "Yarn-Mistral-7b-128k",
maxTokens = 16384,
provider = APIProvider.ModelsLab,
inputTokenPricePerK = 0.0005,
outputTokenPricePerK = 0.0015
)
/*
val PygmalionAI = ChatModels(
name = "PygmalionAI",
modelName = "PygmalionAI",
maxTokens = 16384,
provider = APIProvider.ModelsLab,
inputTokenPricePerK = 0.0005,
outputTokenPricePerK = 0.0015
)
*/
private val Pygmalion13b = ChatModels(
name = "Pygmalion13b",
modelName = "pygmalion-1.3b",
maxTokens = 16384,
provider = APIProvider.ModelsLab,
inputTokenPricePerK = 0.0005,
outputTokenPricePerK = 0.0015
)
private val Opt67b = ChatModels(
name = "Opt67b",
modelName = "opt-6.7b",
maxTokens = 16384,
provider = APIProvider.ModelsLab,
inputTokenPricePerK = 0.0005,
outputTokenPricePerK = 0.0015
)
private val MistralLite = ChatModels(
name = "MistralLite",
modelName = "MistralLite",
maxTokens = 16384,
provider = APIProvider.ModelsLab,
inputTokenPricePerK = 0.0005,
outputTokenPricePerK = 0.0015
)
private val Openchat35 = ChatModels(
name = "Openchat35",
modelName = "openchat_3.5",
maxTokens = 16384,
provider = APIProvider.ModelsLab,
inputTokenPricePerK = 0.0005,
outputTokenPricePerK = 0.0015
)
private val NeuralChat7bV3 = ChatModels(
name = "NeuralChat7bV3",
modelName = "neural-chat-7b-v3",
maxTokens = 16384,
provider = APIProvider.ModelsLab,
inputTokenPricePerK = 0.0005,
outputTokenPricePerK = 0.0015
)
private val OpenHermes25Mistral7B = ChatModels(
name = "OpenHermes25Mistral7B",
modelName = "OpenHermes-2.5-Mistral-7B",
maxTokens = 16384,
provider = APIProvider.ModelsLab,
inputTokenPricePerK = 0.0005,
outputTokenPricePerK = 0.0015
)
private val Dolphin221Mistral7b = ChatModels(
name = "Dolphin221Mistral7b",
modelName = "dolphin-2.2.1-mistral-7b",
maxTokens = 16384,
provider = APIProvider.ModelsLab,
inputTokenPricePerK = 0.0005,
outputTokenPricePerK = 0.0015
)
private val Mistral7BOpenOrca = ChatModels(
name = "Mistral7BOpenOrca",
modelName = "Mistral-7B-OpenOrca",
maxTokens = 16384,
provider = APIProvider.ModelsLab,
inputTokenPricePerK = 0.0005,
outputTokenPricePerK = 0.0015
)
// val CodeLlama7bHf = ChatModels(
// name = "CodeLlama7bHf",
// modelName = "CodeLlama-7b-hf",
// maxTokens = 16384,
// provider = APIProvider.ModelsLab,
// inputTokenPricePerK = 0.0005,
// outputTokenPricePerK = 0.0015
// )
private val DeepseekCoder67bInstruct = ChatModels(
name = "DeepseekCoder67bInstruct",
modelName = "deepseek-coder-6.7b-instruct",
maxTokens = 16384,
provider = APIProvider.ModelsLab,
inputTokenPricePerK = 0.0005,
outputTokenPricePerK = 0.0015
)
private val Phi15 = ChatModels(
name = "Phi15",
modelName = "phi-1_5",
maxTokens = 16384,
provider = APIProvider.ModelsLab,
inputTokenPricePerK = 0.0005,
outputTokenPricePerK = 0.0015
)
private val Zephyr7bAlpha = ChatModels(
name = "Zephyr7bAlpha",
modelName = "zephyr-7b-alpha",
maxTokens = 16384,
provider = APIProvider.ModelsLab,
inputTokenPricePerK = 0.0005,
outputTokenPricePerK = 0.0015
)
private val AWSLLaMA270bChat = ChatModels(
name = "AWSLLaMA270bChat",
modelName = "meta.llama2-70b-chat-v1",
maxTokens = 2048,
provider = APIProvider.AWS,
inputTokenPricePerK = 0.0005,
outputTokenPricePerK = 0.0015
)
private val AWSLLaMA213bChat = ChatModels(
name = "AWSLLaMA213bChat",
modelName = "meta.llama2-13b-chat-v1",
maxTokens = 2048,
provider = APIProvider.AWS,
inputTokenPricePerK = 0.0005,
outputTokenPricePerK = 0.0015
)
private val Mistral7bInstructV02 = ChatModels(
name = "Mistral7bInstructV02",
modelName = "mistral.mistral-7b-instruct-v0:2",
maxTokens = 2 * 1024,
provider = APIProvider.AWS,
inputTokenPricePerK = 0.0005,
outputTokenPricePerK = 0.0015
)
private val Mixtral8x7bInstructV01AWS = ChatModels(
name = "Mixtral8x7bInstructV01AWS",
modelName = "mistral.mixtral-8x7b-instruct-v0:1",
maxTokens = 2 * 1024,
provider = APIProvider.AWS,
inputTokenPricePerK = 0.0005,
outputTokenPricePerK = 0.0015
)
// val AmazonTitanEmbedTextV1 = ChatModels(
// name = "AmazonTitanEmbedTextV1",
// modelName = "amazon.titan-embed-text-v1",
// maxTokens = 16384,
// provider = APIProvider.AWS,
// inputTokenPricePerK = 0.0005,
// outputTokenPricePerK = 0.0015
// )
private val AmazonTitanTextLiteV1 = ChatModels(
name = "AmazonTitanTextLiteV1",
modelName = "amazon.titan-text-lite-v1",
maxTokens = 4096,
provider = APIProvider.AWS,
inputTokenPricePerK = 0.0005,
outputTokenPricePerK = 0.0015
)
private val AmazonTitanTextExpressV1 = ChatModels(
name = "AmazonTitanTextExpressV1",
modelName = "amazon.titan-text-express-v1",
maxTokens = 8192,
provider = APIProvider.AWS,
inputTokenPricePerK = 0.0005,
outputTokenPricePerK = 0.0015
)
private val CohereCommandTextV14 = ChatModels(
name = "CohereCommandTextV14",
modelName = "cohere.command-text-v14",
maxTokens = 4000,
provider = APIProvider.AWS,
inputTokenPricePerK = 0.0005,
outputTokenPricePerK = 0.0015
)
private val AI21J2UltraV1 = ChatModels(
name = "AI21J2UltraV1",
modelName = "ai21.j2-ultra-v1",
maxTokens = 8191,
provider = APIProvider.AWS,
inputTokenPricePerK = 0.0005,
outputTokenPricePerK = 0.0015
)
private val AI21J2MidV1 = ChatModels(
name = "AI21J2MidV1",
modelName = "ai21.j2-mid-v1",
maxTokens = 8191,
provider = APIProvider.AWS,
inputTokenPricePerK = 0.0005,
outputTokenPricePerK = 0.0015
)
private val Claude3Sonnet = ChatModels(
name = "Claude3Sonnet",
modelName = "anthropic.claude-3-sonnet-20240229-v1:0",
maxTokens = 200000,
provider = APIProvider.AWS,
inputTokenPricePerK = 0.0005,
outputTokenPricePerK = 0.0015
)
private val Claude3Haiku = ChatModels(
name = "Claude3Haiku",
modelName = "anthropic.claude-3-haiku-20240307-v1:0",
maxTokens = 200000,
provider = APIProvider.AWS,
inputTokenPricePerK = 0.0005,
outputTokenPricePerK = 0.0015
)
private val ClaudeV2_1 = ChatModels(
name = "ClaudeV2",
modelName = "anthropic.claude-v2:1",
maxTokens = 100000,
provider = APIProvider.AWS,
inputTokenPricePerK = 0.0005,
outputTokenPricePerK = 0.0015
)
private val ClaudeV2 = ChatModels(
name = "ClaudeV2",
modelName = "anthropic.claude-v2",
maxTokens = 100000,
provider = APIProvider.AWS,
inputTokenPricePerK = 0.0005,
outputTokenPricePerK = 0.0015
)
private val ClaudeV2Instant = ChatModels(
name = "ClaudeV2",
modelName = "anthropic.claude-instant-v1",
maxTokens = 100000,
provider = APIProvider.AWS,
inputTokenPricePerK = 0.0005,
outputTokenPricePerK = 0.0015
)
fun values() = mapOf(
"GPT35Turbo" to GPT35Turbo,
// "GPT4" to GPT4,
"GPT4Turbo" to GPT4Turbo,
"GPT4Vision" to GPT4Vision,
"SonarSmallChat" to SonarSmallChat,
"SonarSmallOnline" to SonarSmallOnline,
"SonarMediumChat" to SonarMediumChat,
"SonarMediumOnline" to SonarMediumOnline,
"Codellama70bInstruct" to Codellama70bInstruct,
"Mistral7bInstruct" to Mistral7bInstruct,
"Mixtral8x7bInstruct" to Mixtral8x7bInstruct,
"LLaMA270bChat" to LLaMA270bChat,
"Mixtral8x7bInstructV01" to Mixtral8x7bInstructV01,
// "Gemma7bIt" to Gemma7bIt,
"Zephyr7bBeta" to Zephyr7bBeta,
"DialoGPTLarge" to DialoGPTLarge,
"YarnMistral7b128k" to YarnMistral7b128k,
// "PygmalionAI" to PygmalionAI,
"Pygmalion13b" to Pygmalion13b,
"Opt67b" to Opt67b,
"MistralLite" to MistralLite,
"Openchat35" to Openchat35,
"NeuralChat7bV3" to NeuralChat7bV3,
"OpenHermes25Mistral7B" to OpenHermes25Mistral7B,
"Dolphin221Mistral7b" to Dolphin221Mistral7b,
"Mistral7BOpenOrca" to Mistral7BOpenOrca,
// "CodeLlama7bHf" to CodeLlama7bHf,
"DeepseekCoder67bInstruct" to DeepseekCoder67bInstruct,
"Phi15" to Phi15,
"Zephyr7bAlpha" to Zephyr7bAlpha,
"AWSLLaMA270bChat" to AWSLLaMA270bChat,
"AWSLLaMA213bChat" to AWSLLaMA213bChat,
"Mistral7bInstructV02" to Mistral7bInstructV02,
"Mixtral8x7bInstructV01AWS" to Mixtral8x7bInstructV01AWS,
// "AmazonTitanEmbedTextV1" to AmazonTitanEmbedTextV1,
"AmazonTitanTextLiteV1" to AmazonTitanTextLiteV1,
"AmazonTitanTextExpressV1" to AmazonTitanTextExpressV1,
"CohereCommandTextV14" to CohereCommandTextV14,
"AI21J2UltraV1" to AI21J2UltraV1,
"AI21J2MidV1" to AI21J2MidV1,
"Claude3Sonnet" to Claude3Sonnet,
"Claude3Haiku" to Claude3Haiku,
"ClaudeV2_1" to ClaudeV2_1,
"ClaudeV2" to ClaudeV2,
"ClaudeV2Instant" to ClaudeV2Instant,
)
}
}
class ChatModelsSerializer : StdSerializer(ChatModels::class.java) {
override fun serialize(value: ChatModels, gen: JsonGenerator, provider: SerializerProvider) {
values().entries.find { it.value == value }?.key?.let { gen.writeString(it) }
}
}
class ChatModelsDeserializer : JsonDeserializer() {
override fun deserialize(p: JsonParser, ctxt: DeserializationContext): ChatModels {
val modelName = p.readValueAs(String::class.java)
return ChatModels.values()[modelName] ?: throw IllegalArgumentException("Unknown model name: $modelName")
}
}
© 2015 - 2025 Weber Informatics LLC | Privacy Policy