This commit is contained in:
pacnpal
2025-02-03 10:42:22 -05:00
parent 4c4e9447e4
commit 1515c4faf0

View File

@@ -64,15 +64,13 @@ export interface ApiHandlerOptions {
azureAiEndpoint?: string
azureAiKey?: string
azureAiModelConfig?: ModelInfo
azureAiDeployments?:
| {
[key: string]: {
name: string
apiVersion: string
modelMeshName?: string
}
}
| undefined
azureAiDeployments?: {
[key: string]: {
name: string
apiVersion: string
modelMeshName?: string
}
}
}
export type ApiConfiguration = ApiHandlerOptions & {
@@ -81,13 +79,12 @@ export type ApiConfiguration = ApiHandlerOptions & {
}
// Models
export interface ModelInfo {
maxTokens?: number
contextWindow: number
supportsImages?: boolean
supportsComputerUse?: boolean
supportsPromptCache: boolean // this value is hardcoded for now
supportsPromptCache: boolean
inputPrice?: number
outputPrice?: number
cacheWritesPrice?: number
@@ -96,532 +93,17 @@ export interface ModelInfo {
reasoningEffort?: "low" | "medium" | "high"
}
// Anthropic
// https://docs.anthropic.com/en/docs/about-claude/models
export type AnthropicModelId = keyof typeof anthropicModels
export const anthropicDefaultModelId: AnthropicModelId = "claude-3-5-sonnet-20241022"
export const anthropicModels = {
"claude-3-5-sonnet-20241022": {
maxTokens: 8192,
contextWindow: 200_000,
supportsImages: true,
supportsComputerUse: true,
supportsPromptCache: true,
inputPrice: 3.0, // $3 per million input tokens
outputPrice: 15.0, // $15 per million output tokens
cacheWritesPrice: 3.75, // $3.75 per million tokens
cacheReadsPrice: 0.3, // $0.30 per million tokens
},
"claude-3-5-haiku-20241022": {
maxTokens: 8192,
contextWindow: 200_000,
supportsImages: false,
supportsPromptCache: true,
inputPrice: 1.0,
outputPrice: 5.0,
cacheWritesPrice: 1.25,
cacheReadsPrice: 0.1,
},
"claude-3-opus-20240229": {
maxTokens: 4096,
contextWindow: 200_000,
supportsImages: true,
supportsPromptCache: true,
inputPrice: 15.0,
outputPrice: 75.0,
cacheWritesPrice: 18.75,
cacheReadsPrice: 1.5,
},
"claude-3-haiku-20240307": {
maxTokens: 4096,
contextWindow: 200_000,
supportsImages: true,
supportsPromptCache: true,
inputPrice: 0.25,
outputPrice: 1.25,
cacheWritesPrice: 0.3,
cacheReadsPrice: 0.03,
},
} as const satisfies Record<string, ModelInfo> // as const assertion makes the object deeply readonly
// AWS Bedrock
// https://docs.aws.amazon.com/bedrock/latest/userguide/conversation-inference.html
export interface MessageContent {
type: "text" | "image" | "video" | "tool_use" | "tool_result"
text?: string
source?: {
type: "base64"
data: string | Uint8Array // string for Anthropic, Uint8Array for Bedrock
media_type: "image/jpeg" | "image/png" | "image/gif" | "image/webp"
}
// Video specific fields
format?: string
s3Location?: {
uri: string
bucketOwner?: string
}
// Tool use and result fields
toolUseId?: string
name?: string
input?: any
output?: any // Used for tool_result type
// Azure AI Model Inference Configuration
export interface AzureDeploymentConfig {
name: string
apiVersion: string // Azure AI Inference API version (e.g. 2024-05-01-preview)
modelMeshName?: string // Model-Mesh deployment name if using Model-Mesh
}
export type BedrockModelId = keyof typeof bedrockModels
export const bedrockDefaultModelId: BedrockModelId = "anthropic.claude-3-5-sonnet-20241022-v2:0"
export const bedrockModels = {
"amazon.nova-pro-v1:0": {
maxTokens: 5000,
contextWindow: 300_000,
supportsImages: true,
supportsComputerUse: false,
supportsPromptCache: false,
inputPrice: 0.8,
outputPrice: 3.2,
cacheWritesPrice: 0.8, // per million tokens
cacheReadsPrice: 0.2, // per million tokens
},
"amazon.nova-lite-v1:0": {
maxTokens: 5000,
contextWindow: 300_000,
supportsImages: true,
supportsComputerUse: false,
supportsPromptCache: false,
inputPrice: 0.06,
outputPrice: 0.024,
cacheWritesPrice: 0.06, // per million tokens
cacheReadsPrice: 0.015, // per million tokens
},
"amazon.nova-micro-v1:0": {
maxTokens: 5000,
contextWindow: 128_000,
supportsImages: false,
supportsComputerUse: false,
supportsPromptCache: false,
inputPrice: 0.035,
outputPrice: 0.14,
cacheWritesPrice: 0.035, // per million tokens
cacheReadsPrice: 0.00875, // per million tokens
},
"anthropic.claude-3-5-sonnet-20241022-v2:0": {
maxTokens: 8192,
contextWindow: 200_000,
supportsImages: true,
supportsComputerUse: true,
supportsPromptCache: false,
inputPrice: 3.0,
outputPrice: 15.0,
cacheWritesPrice: 3.75, // per million tokens
cacheReadsPrice: 0.3, // per million tokens
},
"anthropic.claude-3-5-haiku-20241022-v1:0": {
maxTokens: 8192,
contextWindow: 200_000,
supportsImages: false,
supportsPromptCache: false,
inputPrice: 1.0,
outputPrice: 5.0,
cacheWritesPrice: 1.0,
cacheReadsPrice: 0.08,
},
"anthropic.claude-3-5-sonnet-20240620-v1:0": {
maxTokens: 8192,
contextWindow: 200_000,
supportsImages: true,
supportsPromptCache: false,
inputPrice: 3.0,
outputPrice: 15.0,
},
"anthropic.claude-3-opus-20240229-v1:0": {
maxTokens: 4096,
contextWindow: 200_000,
supportsImages: true,
supportsPromptCache: false,
inputPrice: 15.0,
outputPrice: 75.0,
},
"anthropic.claude-3-sonnet-20240229-v1:0": {
maxTokens: 4096,
contextWindow: 200_000,
supportsImages: true,
supportsPromptCache: false,
inputPrice: 3.0,
outputPrice: 15.0,
},
"anthropic.claude-3-haiku-20240307-v1:0": {
maxTokens: 4096,
contextWindow: 200_000,
supportsImages: true,
supportsPromptCache: false,
inputPrice: 0.25,
outputPrice: 1.25,
},
"meta.llama3-3-70b-instruct-v1:0": {
maxTokens: 8192,
contextWindow: 128_000,
supportsImages: false,
supportsComputerUse: false,
supportsPromptCache: false,
inputPrice: 0.72,
outputPrice: 0.72,
},
"meta.llama3-2-90b-instruct-v1:0": {
maxTokens: 8192,
contextWindow: 128_000,
supportsImages: true,
supportsComputerUse: false,
supportsPromptCache: false,
inputPrice: 0.72,
outputPrice: 0.72,
},
"meta.llama3-2-11b-instruct-v1:0": {
maxTokens: 8192,
contextWindow: 128_000,
supportsImages: true,
supportsComputerUse: false,
supportsPromptCache: false,
inputPrice: 0.16,
outputPrice: 0.16,
},
"meta.llama3-2-3b-instruct-v1:0": {
maxTokens: 8192,
contextWindow: 128_000,
supportsImages: false,
supportsComputerUse: false,
supportsPromptCache: false,
inputPrice: 0.15,
outputPrice: 0.15,
},
"meta.llama3-2-1b-instruct-v1:0": {
maxTokens: 8192,
contextWindow: 128_000,
supportsImages: false,
supportsComputerUse: false,
supportsPromptCache: false,
inputPrice: 0.1,
outputPrice: 0.1,
},
"meta.llama3-1-405b-instruct-v1:0": {
maxTokens: 8192,
contextWindow: 128_000,
supportsImages: false,
supportsComputerUse: false,
supportsPromptCache: false,
inputPrice: 2.4,
outputPrice: 2.4,
},
"meta.llama3-1-70b-instruct-v1:0": {
maxTokens: 8192,
contextWindow: 128_000,
supportsImages: false,
supportsComputerUse: false,
supportsPromptCache: false,
inputPrice: 0.72,
outputPrice: 0.72,
},
"meta.llama3-1-8b-instruct-v1:0": {
maxTokens: 8192,
contextWindow: 8_000,
supportsImages: false,
supportsComputerUse: false,
supportsPromptCache: false,
inputPrice: 0.22,
outputPrice: 0.22,
},
"meta.llama3-70b-instruct-v1:0": {
maxTokens: 2048,
contextWindow: 8_000,
supportsImages: false,
supportsComputerUse: false,
supportsPromptCache: false,
inputPrice: 2.65,
outputPrice: 3.5,
},
"meta.llama3-8b-instruct-v1:0": {
maxTokens: 2048,
contextWindow: 4_000,
supportsImages: false,
supportsComputerUse: false,
supportsPromptCache: false,
inputPrice: 0.3,
outputPrice: 0.6,
},
} as const satisfies Record<string, ModelInfo>
// Glama
// https://glama.ai/models
export const glamaDefaultModelId = "anthropic/claude-3-5-sonnet"
export const glamaDefaultModelInfo: ModelInfo = {
maxTokens: 8192,
contextWindow: 200_000,
supportsImages: true,
supportsComputerUse: true,
supportsPromptCache: true,
inputPrice: 3.0,
outputPrice: 15.0,
cacheWritesPrice: 3.75,
cacheReadsPrice: 0.3,
description:
"The new Claude 3.5 Sonnet delivers better-than-Opus capabilities, faster-than-Sonnet speeds, at the same Sonnet prices. Sonnet is particularly good at:\n\n- Coding: New Sonnet scores ~49% on SWE-Bench Verified, higher than the last best score, and without any fancy prompt scaffolding\n- Data science: Augments human data science expertise; navigates unstructured data while using multiple tools for insights\n- Visual processing: excelling at interpreting charts, graphs, and images, accurately transcribing text to derive insights beyond just the text alone\n- Agentic tasks: exceptional tool use, making it great at agentic tasks (i.e. complex, multi-step problem solving tasks that require engaging with other systems)\n\n#multimodal\n\n_This is a faster endpoint, made available in collaboration with Anthropic, that is self-moderated: response moderation happens on the provider's side instead of OpenRouter's. For requests that pass moderation, it's identical to the [Standard](/anthropic/claude-3.5-sonnet) variant._",
}
// OpenRouter
// https://openrouter.ai/models?order=newest&supported_parameters=tools
export const openRouterDefaultModelId = "anthropic/claude-3.5-sonnet:beta" // will always exist in openRouterModels
export const openRouterDefaultModelInfo: ModelInfo = {
maxTokens: 8192,
contextWindow: 200_000,
supportsImages: true,
supportsComputerUse: true,
supportsPromptCache: true,
inputPrice: 3.0,
outputPrice: 15.0,
cacheWritesPrice: 3.75,
cacheReadsPrice: 0.3,
description:
"The new Claude 3.5 Sonnet delivers better-than-Opus capabilities, faster-than-Sonnet speeds, at the same Sonnet prices. Sonnet is particularly good at:\n\n- Coding: New Sonnet scores ~49% on SWE-Bench Verified, higher than the last best score, and without any fancy prompt scaffolding\n- Data science: Augments human data science expertise; navigates unstructured data while using multiple tools for insights\n- Visual processing: excelling at interpreting charts, graphs, and images, accurately transcribing text to derive insights beyond just the text alone\n- Agentic tasks: exceptional tool use, making it great at agentic tasks (i.e. complex, multi-step problem solving tasks that require engaging with other systems)\n\n#multimodal\n\n_This is a faster endpoint, made available in collaboration with Anthropic, that is self-moderated: response moderation happens on the provider's side instead of OpenRouter's. For requests that pass moderation, it's identical to the [Standard](/anthropic/claude-3.5-sonnet) variant._",
}
// Vertex AI
// https://cloud.google.com/vertex-ai/generative-ai/docs/partner-models/use-claude
export type VertexModelId = keyof typeof vertexModels
export const vertexDefaultModelId: VertexModelId = "claude-3-5-sonnet-v2@20241022"
export const vertexModels = {
"claude-3-5-sonnet-v2@20241022": {
maxTokens: 8192,
contextWindow: 200_000,
supportsImages: true,
supportsComputerUse: true,
supportsPromptCache: false,
inputPrice: 3.0,
outputPrice: 15.0,
},
"claude-3-5-sonnet@20240620": {
maxTokens: 8192,
contextWindow: 200_000,
supportsImages: true,
supportsPromptCache: false,
inputPrice: 3.0,
outputPrice: 15.0,
},
"claude-3-5-haiku@20241022": {
maxTokens: 8192,
contextWindow: 200_000,
supportsImages: false,
supportsPromptCache: false,
inputPrice: 1.0,
outputPrice: 5.0,
},
"claude-3-opus@20240229": {
maxTokens: 4096,
contextWindow: 200_000,
supportsImages: true,
supportsPromptCache: false,
inputPrice: 15.0,
outputPrice: 75.0,
},
"claude-3-haiku@20240307": {
maxTokens: 4096,
contextWindow: 200_000,
supportsImages: true,
supportsPromptCache: false,
inputPrice: 0.25,
outputPrice: 1.25,
},
} as const satisfies Record<string, ModelInfo>
export const openAiModelInfoSaneDefaults: ModelInfo = {
maxTokens: -1,
contextWindow: 128_000,
supportsImages: true,
supportsPromptCache: false,
inputPrice: 0,
outputPrice: 0,
}
// Gemini
// https://ai.google.dev/gemini-api/docs/models/gemini
export type GeminiModelId = keyof typeof geminiModels
export const geminiDefaultModelId: GeminiModelId = "gemini-2.0-flash-thinking-exp-01-21"
export const geminiModels = {
"gemini-2.0-flash-thinking-exp-01-21": {
maxTokens: 65_536,
contextWindow: 1_048_576,
supportsImages: true,
supportsPromptCache: false,
inputPrice: 0,
outputPrice: 0,
},
"gemini-2.0-flash-thinking-exp-1219": {
maxTokens: 8192,
contextWindow: 32_767,
supportsImages: true,
supportsPromptCache: false,
inputPrice: 0,
outputPrice: 0,
},
"gemini-2.0-flash-exp": {
maxTokens: 8192,
contextWindow: 1_048_576,
supportsImages: true,
supportsPromptCache: false,
inputPrice: 0,
outputPrice: 0,
},
"gemini-1.5-flash-002": {
maxTokens: 8192,
contextWindow: 1_048_576,
supportsImages: true,
supportsPromptCache: false,
inputPrice: 0,
outputPrice: 0,
},
"gemini-1.5-flash-exp-0827": {
maxTokens: 8192,
contextWindow: 1_048_576,
supportsImages: true,
supportsPromptCache: false,
inputPrice: 0,
outputPrice: 0,
},
"gemini-1.5-flash-8b-exp-0827": {
maxTokens: 8192,
contextWindow: 1_048_576,
supportsImages: true,
supportsPromptCache: false,
inputPrice: 0,
outputPrice: 0,
},
"gemini-1.5-pro-002": {
maxTokens: 8192,
contextWindow: 2_097_152,
supportsImages: true,
supportsPromptCache: false,
inputPrice: 0,
outputPrice: 0,
},
"gemini-1.5-pro-exp-0827": {
maxTokens: 8192,
contextWindow: 2_097_152,
supportsImages: true,
supportsPromptCache: false,
inputPrice: 0,
outputPrice: 0,
},
"gemini-exp-1206": {
maxTokens: 8192,
contextWindow: 2_097_152,
supportsImages: true,
supportsPromptCache: false,
inputPrice: 0,
outputPrice: 0,
},
} as const satisfies Record<string, ModelInfo>
// OpenAI Native
// https://openai.com/api/pricing/
export type OpenAiNativeModelId = keyof typeof openAiNativeModels
export const openAiNativeDefaultModelId: OpenAiNativeModelId = "gpt-4o"
export const openAiNativeModels = {
// don't support tool use yet
"o3-mini": {
maxTokens: 100_000,
contextWindow: 200_000,
supportsImages: false,
supportsPromptCache: false,
inputPrice: 1.1,
outputPrice: 4.4,
reasoningEffort: "medium",
},
"o3-mini-high": {
maxTokens: 100_000,
contextWindow: 200_000,
supportsImages: false,
supportsPromptCache: false,
inputPrice: 1.1,
outputPrice: 4.4,
reasoningEffort: "high",
},
"o3-mini-low": {
maxTokens: 100_000,
contextWindow: 200_000,
supportsImages: false,
supportsPromptCache: false,
inputPrice: 1.1,
outputPrice: 4.4,
reasoningEffort: "low",
},
o1: {
maxTokens: 100_000,
contextWindow: 200_000,
supportsImages: true,
supportsPromptCache: false,
inputPrice: 15,
outputPrice: 60,
},
"o1-preview": {
maxTokens: 32_768,
contextWindow: 128_000,
supportsImages: true,
supportsPromptCache: false,
inputPrice: 15,
outputPrice: 60,
},
"o1-mini": {
maxTokens: 65_536,
contextWindow: 128_000,
supportsImages: true,
supportsPromptCache: false,
inputPrice: 1.1,
outputPrice: 4.4,
},
"gpt-4o": {
maxTokens: 4_096,
contextWindow: 128_000,
supportsImages: true,
supportsPromptCache: false,
inputPrice: 5,
outputPrice: 15,
},
"gpt-4o-mini": {
maxTokens: 16_384,
contextWindow: 128_000,
supportsImages: true,
supportsPromptCache: false,
inputPrice: 0.15,
outputPrice: 0.6,
},
} as const satisfies Record<string, ModelInfo>
// DeepSeek
// https://platform.deepseek.com/docs/api
export type DeepSeekModelId = keyof typeof deepSeekModels
export const deepSeekDefaultModelId: DeepSeekModelId = "deepseek-chat"
export const deepSeekModels = {
"deepseek-chat": {
maxTokens: 8192,
contextWindow: 64_000,
supportsImages: false,
supportsPromptCache: false,
inputPrice: 0.014, // $0.014 per million tokens
outputPrice: 0.28, // $0.28 per million tokens
description: `DeepSeek-V3 achieves a significant breakthrough in inference speed over previous models. It tops the leaderboard among open-source models and rivals the most advanced closed-source models globally.`,
},
"deepseek-reasoner": {
maxTokens: 8192,
contextWindow: 64_000,
supportsImages: false,
supportsPromptCache: false,
inputPrice: 0.55, // $0.55 per million tokens
outputPrice: 2.19, // $2.19 per million tokens
description: `DeepSeek-R1 achieves performance comparable to OpenAI-o1 across math, code, and reasoning tasks.`,
},
} as const satisfies Record<string, ModelInfo>
// Azure OpenAI
// https://learn.microsoft.com/en-us/azure/ai-services/openai/api-version-deprecation
// https://learn.microsoft.com/en-us/azure/ai-services/openai/reference#api-specs
// Azure OpenAI API Version
export const azureOpenAiDefaultApiVersion = "2024-08-01-preview"
// Azure AI Model Inference Defaults
export const azureAiModelInfoSaneDefaults: ModelInfo = {
maxTokens: -1, // Dynamic based on model
contextWindow: 128_000, // Conservative default
@@ -632,77 +114,3 @@ export const azureAiModelInfoSaneDefaults: ModelInfo = {
outputPrice: 0,
description: "Azure AI Model Inference allows you to deploy and use any model through Azure's inference service.",
}
// Mistral
// https://docs.mistral.ai/getting-started/models/models_overview/
export type MistralModelId = keyof typeof mistralModels
export const mistralDefaultModelId: MistralModelId = "codestral-latest"
export const mistralModels = {
"codestral-latest": {
maxTokens: 32_768,
contextWindow: 256_000,
supportsImages: false,
supportsPromptCache: false,
inputPrice: 0.3,
outputPrice: 0.9,
},
} as const satisfies Record<string, ModelInfo>
// Unbound Security
export type UnboundModelId = keyof typeof unboundModels
export const unboundDefaultModelId = "openai/gpt-4o"
export const unboundModels = {
"anthropic/claude-3-5-sonnet-20241022": anthropicModels["claude-3-5-sonnet-20241022"],
"openai/gpt-4o": openAiNativeModels["gpt-4o"],
"deepseek/deepseek-chat": deepSeekModels["deepseek-chat"],
"deepseek/deepseek-reasoner": deepSeekModels["deepseek-reasoner"],
"mistral/codestral-latest": mistralModels["codestral-latest"],
} as const satisfies Record<string, ModelInfo>
// Azure AI Model Inference Configuration
export interface AzureDeploymentConfig {
name: string
apiVersion: string
modelMeshName?: string // For Model-Mesh deployments
}
// Azure OpenAI Models
export type AzureAiModelId = "azure-gpt-35" | "azure-gpt-4" | "azure-gpt-4-turbo"
export const azureAiModels: Record<AzureAiModelId, ModelInfo & { defaultDeployment: AzureDeploymentConfig }> = {
"azure-gpt-35": {
maxTokens: 4096,
contextWindow: 16385,
supportsPromptCache: true,
inputPrice: 0.0015,
outputPrice: 0.002,
defaultDeployment: {
name: "azure-gpt-35",
apiVersion: "2024-02-15-preview",
},
},
"azure-gpt-4": {
maxTokens: 8192,
contextWindow: 8192,
supportsPromptCache: true,
inputPrice: 0.03,
outputPrice: 0.06,
defaultDeployment: {
name: "azure-gpt-4",
apiVersion: "2024-02-15-preview",
},
},
"azure-gpt-4-turbo": {
maxTokens: 4096,
contextWindow: 128000,
supportsPromptCache: true,
inputPrice: 0.01,
outputPrice: 0.03,
defaultDeployment: {
name: "azure-gpt-4-turbo",
apiVersion: "2024-02-15-preview",
},
},
} as const satisfies Record<AzureAiModelId, ModelInfo & { defaultDeployment: AzureDeploymentConfig }>
export const azureAiDefaultModelId: AzureAiModelId = "azure-gpt-35"