@llmgateway/ai-sdk-provider 1.0.0

package/dist/index.d.ts

@@ -0,0 +1,263 @@
+import { LanguageModelV1 } from '@ai-sdk/provider';
+export { LanguageModelV1 } from '@ai-sdk/provider';
+
+type LLMGatewayLanguageModel = LanguageModelV1;
+type LLMGatewayProviderOptions = {
+    models?: string[];
+    /**
+     * https://llmgateway.io/docs/use-cases/reasoning-tokens
+     * One of `max_tokens` or `effort` is required.
+     * If `exclude` is true, reasoning will be removed from the response. Default is false.
+     */
+    reasoning?: {
+        exclude?: boolean;
+    } & ({
+        max_tokens: number;
+    } | {
+        effort: 'high' | 'medium' | 'low';
+    });
+    /**
+     * A unique identifier representing your end-user, which can
+     * help LLMGateway to monitor and detect abuse.
+     */
+    user?: string;
+};
+type LLMGatewaySharedSettings = LLMGatewayProviderOptions & {
+    /**
+     * @deprecated use `reasoning` instead
+     */
+    includeReasoning?: boolean;
+    extraBody?: Record<string, unknown>;
+    /**
+     * Enable usage accounting to get detailed token usage information.
+     * https://llmgateway.io/docs/use-cases/usage-accounting
+     */
+    usage?: {
+        /**
+         * When true, includes token usage information in the response.
+         */
+        include: boolean;
+    };
+};
+/**
+ * Usage accounting response
+ * @see https://llmgateway.io/docs/use-cases/usage-accounting
+ */
+type LLMGatewayUsageAccounting = {
+    promptTokens: number;
+    promptTokensDetails?: {
+        cachedTokens: number;
+    };
+    completionTokens: number;
+    completionTokensDetails?: {
+        reasoningTokens: number;
+    };
+    totalTokens: number;
+    cost?: number;
+};
+
+type LLMGatewayCompletionModelId = string;
+type LLMGatewayCompletionSettings = {
+    /**
+  Modify the likelihood of specified tokens appearing in the completion.
+  
+  Accepts a JSON object that maps tokens (specified by their token ID in
+  the GPT tokenizer) to an associated bias value from -100 to 100. You
+  can use this tokenizer tool to convert text to token IDs. Mathematically,
+  the bias is added to the logits generated by the model prior to sampling.
+  The exact effect will vary per model, but values between -1 and 1 should
+  decrease or increase likelihood of selection; values like -100 or 100
+  should result in a ban or exclusive selection of the relevant token.
+  
+  As an example, you can pass {"50256": -100} to prevent the <|endoftext|>
+  token from being generated.
+     */
+    logitBias?: Record<number, number>;
+    /**
+  Return the log probabilities of the tokens. Including logprobs will increase
+  the response size and can slow down response times. However, it can
+  be useful to better understand how the model is behaving.
+  
+  Setting to true will return the log probabilities of the tokens that
+  were generated.
+  
+  Setting to a number will return the log probabilities of the top n
+  tokens that were generated.
+     */
+    logprobs?: boolean | number;
+    /**
+  The suffix that comes after a completion of inserted text.
+     */
+    suffix?: string;
+} & LLMGatewaySharedSettings;
+
+type LLMGatewayChatModelId = string;
+type LLMGatewayChatSettings = {
+    /**
+  Modify the likelihood of specified tokens appearing in the completion.
+  
+  Accepts a JSON object that maps tokens (specified by their token ID in
+  the GPT tokenizer) to an associated bias value from -100 to 100. You
+  can use this tokenizer tool to convert text to token IDs. Mathematically,
+  the bias is added to the logits generated by the model prior to sampling.
+  The exact effect will vary per model, but values between -1 and 1 should
+  decrease or increase likelihood of selection; values like -100 or 100
+  should result in a ban or exclusive selection of the relevant token.
+  
+  As an example, you can pass {"50256": -100} to prevent the <|endoftext|>
+  token from being generated.
+  */
+    logitBias?: Record<number, number>;
+    /**
+  Return the log probabilities of the tokens. Including logprobs will increase
+  the response size and can slow down response times. However, it can
+  be useful to better understand how the model is behaving.
+  
+  Setting to true will return the log probabilities of the tokens that
+  were generated.
+  
+  Setting to a number will return the log probabilities of the top n
+  tokens that were generated.
+  */
+    logprobs?: boolean | number;
+    /**
+  Whether to enable parallel function calling during tool use. Default to true.
+     */
+    parallelToolCalls?: boolean;
+    /**
+  A unique identifier representing your end-user, which can help LLMGateway to
+  monitor and detect abuse. Learn more.
+  */
+    user?: string;
+} & LLMGatewaySharedSettings;
+
+type LLMGatewayChatConfig = {
+    provider: string;
+    compatibility: 'strict' | 'compatible';
+    headers: () => Record<string, string | undefined>;
+    url: (options: {
+        modelId: string;
+        path: string;
+    }) => string;
+    fetch?: typeof fetch;
+    extraBody?: Record<string, unknown>;
+};
+type DoGenerateOutput = Awaited<ReturnType<LanguageModelV1['doGenerate']>>;
+type DoStreamOutput = Awaited<ReturnType<LanguageModelV1['doStream']>>;
+declare class LLMGatewayChatLanguageModel implements LanguageModelV1 {
+    readonly specificationVersion = "v1";
+    readonly defaultObjectGenerationMode = "tool";
+    readonly modelId: LLMGatewayChatModelId;
+    readonly settings: LLMGatewayChatSettings;
+    private readonly config;
+    constructor(modelId: LLMGatewayChatModelId, settings: LLMGatewayChatSettings, config: LLMGatewayChatConfig);
+    get provider(): string;
+    private getArgs;
+    doGenerate(options: Parameters<LanguageModelV1['doGenerate']>[0]): Promise<DoGenerateOutput>;
+    doStream(options: Parameters<LanguageModelV1['doStream']>[0]): Promise<DoStreamOutput>;
+}
+
+type LLMGatewayCompletionConfig = {
+    provider: string;
+    compatibility: 'strict' | 'compatible';
+    headers: () => Record<string, string | undefined>;
+    url: (options: {
+        modelId: string;
+        path: string;
+    }) => string;
+    fetch?: typeof fetch;
+    extraBody?: Record<string, unknown>;
+};
+declare class LLMGatewayCompletionLanguageModel implements LanguageModelV1 {
+    readonly specificationVersion = "v1";
+    readonly defaultObjectGenerationMode: undefined;
+    readonly modelId: LLMGatewayCompletionModelId;
+    readonly settings: LLMGatewayCompletionSettings;
+    private readonly config;
+    constructor(modelId: LLMGatewayCompletionModelId, settings: LLMGatewayCompletionSettings, config: LLMGatewayCompletionConfig);
+    get provider(): string;
+    private getArgs;
+    doGenerate(options: Parameters<LanguageModelV1['doGenerate']>[0]): Promise<Awaited<ReturnType<LanguageModelV1['doGenerate']>>>;
+    doStream(options: Parameters<LanguageModelV1['doStream']>[0]): Promise<Awaited<ReturnType<LanguageModelV1['doStream']>>>;
+}
+
+interface LLMGatewayProvider {
+    (modelId: LLMGatewayChatModelId, settings?: LLMGatewayCompletionSettings): LLMGatewayCompletionLanguageModel;
+    (modelId: LLMGatewayChatModelId, settings?: LLMGatewayChatSettings): LLMGatewayChatLanguageModel;
+    languageModel(modelId: LLMGatewayChatModelId, settings?: LLMGatewayCompletionSettings): LLMGatewayCompletionLanguageModel;
+    languageModel(modelId: LLMGatewayChatModelId, settings?: LLMGatewayChatSettings): LLMGatewayChatLanguageModel;
+    /**
+  Creates an LLMGateway chat model for text generation.
+     */
+    chat(modelId: LLMGatewayChatModelId, settings?: LLMGatewayChatSettings): LLMGatewayChatLanguageModel;
+    /**
+  Creates an LLMGateway completion model for text generation.
+     */
+    completion(modelId: LLMGatewayCompletionModelId, settings?: LLMGatewayCompletionSettings): LLMGatewayCompletionLanguageModel;
+}
+interface LLMGatewayProviderSettings {
+    /**
+  Base URL for the LLMGateway API calls.
+       */
+    baseURL?: string;
+    /**
+  API key for authenticating requests.
+       */
+    apiKey?: string;
+    /**
+  Custom headers to include in the requests.
+       */
+    headers?: Record<string, string>;
+    /**
+  LLMGateway compatibility mode. Should be set to `strict` when using the LLMGateway API,
+  and `compatible` when using 3rd party providers. In `compatible` mode, newer
+  information such as streamOptions are not being sent. Defaults to 'compatible'.
+     */
+    compatibility?: 'strict' | 'compatible';
+    /**
+  Custom fetch implementation. You can use it as a middleware to intercept requests,
+  or to provide a custom fetch implementation for e.g. testing.
+      */
+    fetch?: typeof fetch;
+    /**
+  A JSON object to send as the request body to access LLMGateway features & upstream provider features.
+    */
+    extraBody?: Record<string, unknown>;
+}
+/**
+Create an LLMGateway provider instance.
+ */
+declare function createLLMGateway(options?: LLMGatewayProviderSettings): LLMGatewayProvider;
+/**
+Default LLMGateway provider instance. It uses 'strict' compatibility mode.
+ */
+declare const llmgateway: LLMGatewayProvider;
+
+/**
+@deprecated Use `createLLMGateway` instead.
+ */
+declare class LLMGateway {
+    /**
+  Use a different URL prefix for API calls, e.g. to use proxy servers.
+  The default prefix is `https://api.llmgateway.io/v1`.
+     */
+    readonly baseURL: string;
+    /**
+  API key that is being send using the `Authorization` header.
+  It defaults to the `LLMGATEWAY_API_KEY` environment variable.
+   */
+    readonly apiKey?: string;
+    /**
+  Custom headers to include in the requests.
+     */
+    readonly headers?: Record<string, string>;
+    /**
+     * Creates a new LLMGateway provider instance.
+     */
+    constructor(options?: LLMGatewayProviderSettings);
+    private get baseConfig();
+    chat(modelId: LLMGatewayChatModelId, settings?: LLMGatewayChatSettings): LLMGatewayChatLanguageModel;
+    completion(modelId: LLMGatewayCompletionModelId, settings?: LLMGatewayCompletionSettings): LLMGatewayCompletionLanguageModel;
+}
+
+export { LLMGateway, type LLMGatewayCompletionSettings, type LLMGatewayLanguageModel, type LLMGatewayProvider, type LLMGatewayProviderOptions, type LLMGatewayProviderSettings, type LLMGatewaySharedSettings, type LLMGatewayUsageAccounting, createLLMGateway, llmgateway };