import { GenerateLanguageOptions, GenerateLanguageResult } from './generation-types';
export type { LLMProviderName, GenerateLanguageOptions, GenerateLanguageResult, LanguageAttachment, } from './generation-types';
export { convertMarkdownToHTMLEscaped } from './utils/markdown-to-html';
/**
 * ### Generate Language Response
 * Writes a language response that shows human-like understanding of the
 * question and context.
 * - _Requires_: LLM provider, API key, agent name, and context variables.
 * - _Providers_: groq, togetherai, openai, anthropic, xai, google,
 *   perplexity, cloudflare, nvidia
 * - _Agent Templates_: custom local entries defined in AGENT_PROMPTS.
 * - _How it Works_: Language models predict the most likely next token given
 *   a prompt. They represent words as high-dimensional vectors, use
 *   transformer attention across all prior tokens, and sample from the
 *   resulting probability distribution to produce human-like text.
 *
 * @see [Vercel AI SDK generateText docs](https://sdk.vercel.ai/docs/reference/ai-sdk-core/generate-text)
 * @see [Hugging Face tutorials](https://huggingface.co/learn)
 * @see [Illustrated Transformer](https://jalammar.github.io/illustrated-transformer/)
 * @see [Building a Transformer with PyTorch](https://www.datacamp.com/tutorial/building-a-transformer-with-py-torch)
 * @see [LLM training example](https://github.com/vtempest/ai-research-agent/blob/master/packages/neural-net/src/train/predict-next-word.js)
 *
 * @param options - Configuration for the language-model call
 * @returns Resolved response object with `content`, optional `extract`, or `error`
 * @author [Language Model Researchers](https://arc.net/folder/D0472A20-9C20-4D3F-B145-D2865C0A9FEE)
 * @example
 * const response = await writeLanguageResponse({
 *   query: "Explain neural networks",
 *   agent: "question",
 *   provider: "groq",
 *   apiKey: "your-api-key",
 * });
 */
export declare function writeLanguageResponse(options?: GenerateLanguageOptions): Promise<GenerateLanguageResult>;