inceptionai 0.1.0-alpha.2

package/src/resources/chat/chat.ts

@@ -0,0 +1,182 @@
+// File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details.
+
+import { APIResource } from '../../core/resource';
+import * as CompletionsAPI from './completions';
+import { CompletionCreateParams, Completions } from './completions';
+
+export class Chat extends APIResource {
+  completions: CompletionsAPI.Completions = new CompletionsAPI.Completions(this._client);
+}
+
+export interface ChatCompletion {
+  id: string;
+
+  choices: Array<ChatCompletionChoice>;
+
+  created: number;
+
+  model: string;
+
+  object: 'chat.completion';
+
+  usage: ChatCompletionUsage;
+
+  /**
+   * Summary of the model's reasoning process.
+   */
+  reasoning_summary?: ChatCompletionReasoningSummary | null;
+
+  warning?: string | null;
+}
+
+export interface ChatCompletionChoice {
+  /**
+   * The reason the model stopped generating tokens.
+   */
+  finish_reason: 'stop' | 'length' | 'tool_calls' | 'content_filter' | null;
+
+  index: number;
+
+  message: ChatCompletionMessage;
+}
+
+/**
+ * Model for a single chunk in a streaming chat completion response
+ */
+export interface ChatCompletionChunk {
+  id: string;
+
+  choices: Array<ChatCompletionChunkChoice>;
+
+  created: number;
+
+  model: string;
+
+  object: 'chat.completion.chunk';
+
+  /**
+   * Summary of the model's reasoning process.
+   */
+  reasoning_summary?: ChatCompletionReasoningSummary | null;
+}
+
+/**
+ * Represents a choice in a streaming response chunk
+ */
+export interface ChatCompletionChunkChoice {
+  /**
+   * Represents a partial message in a streaming response chunk
+   */
+  delta: ChatCompletionChunkDelta;
+
+  index: number;
+
+  /**
+   * The reason the model stopped generating tokens.
+   */
+  finish_reason?: 'stop' | 'length' | 'tool_calls' | 'content_filter' | null;
+}
+
+/**
+ * Represents a partial message in a streaming response chunk
+ */
+export interface ChatCompletionChunkDelta {
+  /**
+   * The partial content of the message
+   */
+  content?: string | null;
+
+  /**
+   * The role of the message sender, typically only in the first chunk
+   */
+  role?: string | null;
+
+  /**
+   * The tool calls generated by the model in this chunk
+   */
+  tool_calls?: Array<ChatCompletionToolCall> | null;
+}
+
+export interface ChatCompletionFunctionCall {
+  arguments: string;
+
+  name?: string | null;
+}
+
+export interface ChatCompletionMessage {
+  /**
+   * The role of the message sender (system, user, assistant, tool, function)
+   */
+  role: 'system' | 'user' | 'assistant' | 'tool' | 'function';
+
+  /**
+   * The content of the message, can be string or array of content items
+   */
+  content?: string | Array<{ [key: string]: unknown }> | null;
+
+  /**
+   * The ID of the tool call this message is responding to (for role='tool')
+   */
+  tool_call_id?: string | null;
+
+  /**
+   * The tool calls generated by the model
+   */
+  tool_calls?: Array<ChatCompletionToolCall> | null;
+}
+
+/**
+ * Summary of the model's reasoning process.
+ */
+export interface ChatCompletionReasoningSummary {
+  /**
+   * The summarized reasoning
+   */
+  content?: string | null;
+
+  /**
+   * Status of the summary
+   */
+  status?: 'complete' | 'unavailable';
+}
+
+export interface ChatCompletionToolCall {
+  id: string;
+
+  function: ChatCompletionFunctionCall;
+
+  type: 'function';
+
+  index?: number | null;
+}
+
+export interface ChatCompletionUsage {
+  completion_tokens: number;
+
+  prompt_tokens: number;
+
+  total_tokens: number;
+
+  cached_input_tokens?: number;
+
+  reasoning_tokens?: number;
+}
+
+Chat.Completions = Completions;
+
+export declare namespace Chat {
+  export {
+    type ChatCompletion as ChatCompletion,
+    type ChatCompletionChoice as ChatCompletionChoice,
+    type ChatCompletionChunk as ChatCompletionChunk,
+    type ChatCompletionChunkChoice as ChatCompletionChunkChoice,
+    type ChatCompletionChunkDelta as ChatCompletionChunkDelta,
+    type ChatCompletionFunctionCall as ChatCompletionFunctionCall,
+    type ChatCompletionMessage as ChatCompletionMessage,
+    type ChatCompletionReasoningSummary as ChatCompletionReasoningSummary,
+    type ChatCompletionToolCall as ChatCompletionToolCall,
+    type ChatCompletionUsage as ChatCompletionUsage,
+  };
+
+  export { Completions as Completions, type CompletionCreateParams as CompletionCreateParams };
+}

package/src/resources/chat/completions.ts

@@ -0,0 +1,134 @@
+// File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details.
+
+import { APIResource } from '../../core/resource';
+import * as ChatAPI from './chat';
+import { APIPromise } from '../../core/api-promise';
+import { RequestOptions } from '../../internal/request-options';
+
+/**
+ * Chat completion endpoints (OpenAI-compatible).
+ */
+export class Completions extends APIResource {
+  /**
+   * Generate a chat completion from a sequence of messages. Returns a
+   * `ChatCompletion` object, or a server-sent events stream of `ChatCompletionChunk`
+   * deltas when `stream=true`. Supports tool calling, structured output via
+   * `response_format`, and reasoning controls.
+   *
+   * @example
+   * ```ts
+   * const chatCompletion = await client.chat.completions.create(
+   *   {
+   *     messages: [
+   *       {
+   *         role: 'user',
+   *         content: 'What is a diffusion language model?',
+   *       },
+   *     ],
+   *     model: 'mercury-2',
+   *     max_tokens: 256,
+   *     temperature: 0.75,
+   *   },
+   * );
+   * ```
+   */
+  create(params: CompletionCreateParams, options?: RequestOptions): APIPromise<ChatAPI.ChatCompletion> {
+    const { extra_params, ...body } = params;
+    return this._client.post('/v1/chat/completions', {
+      body: { extra_body: extra_params, ...body },
+      ...options,
+    });
+  }
+}
+
+export interface CompletionCreateParams {
+  messages: Array<ChatAPI.ChatCompletionMessage>;
+
+  /**
+   * The model to use for the chat completion
+   */
+  model: string;
+
+  /**
+   * Whether to show the diffusion effect in the streamed response
+   */
+  diffusing?: boolean;
+
+  /**
+   * Extra parameters passed to the model (OpenAI-compat passthrough).
+   */
+  extra_params?: { [key: string]: unknown };
+
+  /**
+   * Maximum number of tokens to generate
+   */
+  max_tokens?: number;
+
+  /**
+   * Enable flag for more realtime workloads that require lower TTFT/TTFAT
+   */
+  realtime?: boolean;
+
+  /**
+   * Control reasoning effort level (e.g., 'instant', 'low', 'medium', 'high')
+   */
+  reasoning_effort?: string;
+
+  /**
+   * Request a summary of the model's reasoning process
+   */
+  reasoning_summary?: boolean;
+
+  /**
+   * Wait for all reasoning summaries to complete before finishing the response
+   */
+  reasoning_summary_wait?: boolean;
+
+  /**
+   * An object specifying the format that the model must output.
+   */
+  response_format?: { [key: string]: unknown };
+
+  /**
+   * Stop sequences
+   */
+  stop?: Array<string>;
+
+  /**
+   * Whether to stream the response
+   */
+  stream?: boolean;
+
+  /**
+   * Options that control streaming behavior.
+   */
+  stream_options?: CompletionCreateParams.StreamOptions;
+
+  /**
+   * Sampling temperature
+   */
+  temperature?: number;
+
+  /**
+   * Tools to use for the chat completion
+   */
+  tools?: Array<{ [key: string]: unknown }>;
+}
+
+export namespace CompletionCreateParams {
+  /**
+   * Options that control streaming behavior.
+   */
+  export interface StreamOptions {
+    /**
+     * If true, an additional chunk is streamed before the `data: [DONE]` message
+     * containing the token usage statistics for the entire request. Inside that chunk,
+     * `choices` is an empty array and the `usage` field is populated.
+     */
+    include_usage?: boolean;
+  }
+}
+
+export declare namespace Completions {
+  export { type CompletionCreateParams as CompletionCreateParams };
+}

package/src/resources/chat/index.ts

@@ -0,0 +1,16 @@
+// File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details.
+
+export {
+  Chat,
+  type ChatCompletion,
+  type ChatCompletionChoice,
+  type ChatCompletionChunk,
+  type ChatCompletionChunkChoice,
+  type ChatCompletionChunkDelta,
+  type ChatCompletionFunctionCall,
+  type ChatCompletionMessage,
+  type ChatCompletionReasoningSummary,
+  type ChatCompletionToolCall,
+  type ChatCompletionUsage,
+} from './chat';
+export { Completions, type CompletionCreateParams } from './completions';

package/src/resources/chat.ts

@@ -0,0 +1,3 @@
+// File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details.
+
+export * from './chat/index';

package/src/resources/edit/completions.ts

@@ -0,0 +1,71 @@
+// File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details.
+
+import { APIResource } from '../../core/resource';
+import * as EditAPI from './edit';
+import { APIPromise } from '../../core/api-promise';
+import { RequestOptions } from '../../internal/request-options';
+
+/**
+ * Code edit completion endpoints.
+ */
+export class Completions extends APIResource {
+  /**
+   * Generate an edit to a code region given the surrounding context. The request
+   * must contain a single user message whose content includes the required edit
+   * prompt tags (e.g. `<|code_to_edit|>`, `<|cursor|>`, `<|current_file_content|>`).
+   * Returns an `EditCompletion` object. Streaming and tool calling are not supported
+   * on this endpoint.
+   *
+   * @example
+   * ```ts
+   * const editCompletion = await client.edit.completions.create(
+   *   {
+   *     messages: [
+   *       {
+   *         role: 'user',
+   *         content:
+   *           '<|recently_viewed_code_snippets|>\n<|/recently_viewed_code_snippets|>\n<|edit_diff_history|>\n<|/edit_diff_history|>\n<|current_file_content|>\ndef greet(name):\n    print("hi")\n<|/current_file_content|>\n<|code_to_edit|>\ndef greet(name):\n    print("hi")<|cursor|>\n<|/code_to_edit|>',
+   *       },
+   *     ],
+   *     model: 'mercury-edit-2',
+   *   },
+   * );
+   * ```
+   */
+  create(body: CompletionCreateParams, options?: RequestOptions): APIPromise<EditAPI.EditCompletion> {
+    return this._client.post('/v1/edit/completions', { body, ...options });
+  }
+}
+
+export interface CompletionCreateParams {
+  messages: Array<EditAPI.EditCompletionMessageParam>;
+
+  /**
+   * The model to use for the edit completion
+   */
+  model: string;
+
+  /**
+   * Maximum number of tokens to generate
+   */
+  max_tokens?: number;
+
+  /**
+   * Presence penalty parameter
+   */
+  presence_penalty?: number;
+
+  /**
+   * Sampling temperature
+   */
+  temperature?: number;
+
+  /**
+   * Float that controls the cumulative probability of the top tokens to consider
+   */
+  top_p?: number;
+}
+
+export declare namespace Completions {
+  export { type CompletionCreateParams as CompletionCreateParams };
+}

package/src/resources/edit/edit.ts

@@ -0,0 +1,104 @@
+// File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details.
+
+import { APIResource } from '../../core/resource';
+import * as CompletionsAPI from './completions';
+import { CompletionCreateParams, Completions } from './completions';
+
+export class Edit extends APIResource {
+  completions: CompletionsAPI.Completions = new CompletionsAPI.Completions(this._client);
+}
+
+export interface EditCompletion {
+  id: string;
+
+  choices: Array<EditCompletionChoice>;
+
+  created: number;
+
+  model: string;
+
+  object: 'edit.completion';
+
+  /**
+   * Usage for edit completions.
+   */
+  usage: EditCompletionUsage;
+
+  warning?: string | null;
+}
+
+/**
+ * Choice for edit completions.
+ */
+export interface EditCompletionChoice {
+  /**
+   * The reason the model stopped generating tokens.
+   */
+  finish_reason: 'stop' | 'length' | 'content_filter' | null;
+
+  index: number;
+
+  /**
+   * Edit response message. The model always responds as the assistant.
+   */
+  message: EditCompletionMessage;
+}
+
+/**
+ * Edit response message. The model always responds as the assistant.
+ */
+export interface EditCompletionMessage {
+  /**
+   * Edit responses are always from the assistant.
+   */
+  role: 'assistant';
+
+  /**
+   * The model's edit output.
+   */
+  content?: string | Array<{ [key: string]: unknown }> | null;
+}
+
+/**
+ * Edit request message. Edit endpoints only accept a single user message.
+ */
+export interface EditCompletionMessageParam {
+  /**
+   * The content of the edit request. Must contain the required edit prompt tags.
+   */
+  content: string | Array<{ [key: string]: unknown }>;
+
+  /**
+   * Edit requests only accept user messages.
+   */
+  role: 'user';
+}
+
+/**
+ * Usage for edit completions.
+ */
+export interface EditCompletionUsage {
+  completion_tokens: number;
+
+  prompt_tokens: number;
+
+  total_tokens: number;
+
+  cached_input_tokens?: number;
+
+  reasoning_tokens?: number;
+}
+
+Edit.Completions = Completions;
+
+export declare namespace Edit {
+  export {
+    type EditCompletion as EditCompletion,
+    type EditCompletionChoice as EditCompletionChoice,
+    type EditCompletionMessage as EditCompletionMessage,
+    type EditCompletionMessageParam as EditCompletionMessageParam,
+    type EditCompletionUsage as EditCompletionUsage,
+  };
+
+  export { Completions as Completions, type CompletionCreateParams as CompletionCreateParams };
+}

package/src/resources/edit/index.ts

@@ -0,0 +1,11 @@
+// File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details.
+
+export { Completions, type CompletionCreateParams } from './completions';
+export {
+  Edit,
+  type EditCompletion,
+  type EditCompletionChoice,
+  type EditCompletionMessage,
+  type EditCompletionMessageParam,
+  type EditCompletionUsage,
+} from './edit';

package/src/resources/edit.ts

@@ -0,0 +1,3 @@
+// File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details.
+
+export * from './edit/index';

package/src/resources/fim/completions.ts

@@ -0,0 +1,112 @@
+// File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details.
+
+import { APIResource } from '../../core/resource';
+import * as FimAPI from './fim';
+import { APIPromise } from '../../core/api-promise';
+import { RequestOptions } from '../../internal/request-options';
+
+/**
+ * Fill-in-the-middle code completion endpoints.
+ */
+export class Completions extends APIResource {
+  /**
+   * Generate a code completion given a `prompt` (prefix) and optional `suffix`.
+   * Designed for IDE-style inline completion. Returns a `FimCompletion` object, or a
+   * server-sent events stream of `FimCompletionChunk` deltas when `stream=true`.
+   * Tool calling and function calling are not supported.
+   *
+   * @example
+   * ```ts
+   * const fimCompletion = await client.fim.completions.create({
+   *   model: 'mercury-edit-2',
+   *   prompt:
+   *     'def fibonacci(n: int) -> int:\n    if n <= 1:\n        return n\n    return ',
+   *   max_tokens: 256,
+   *   suffix: '\n\nprint(fibonacci(10))\n',
+   * });
+   * ```
+   */
+  create(body: CompletionCreateParams, options?: RequestOptions): APIPromise<FimAPI.FimCompletion> {
+    return this._client.post('/v1/fim/completions', { body, ...options });
+  }
+}
+
+export interface CompletionCreateParams {
+  /**
+   * The model to use for the FIM completion
+   */
+  model: string;
+
+  /**
+   * The prompt to complete
+   */
+  prompt: string;
+
+  /**
+   * Frequency penalty parameter
+   */
+  frequency_penalty?: number;
+
+  /**
+   * Maximum number of tokens to generate
+   */
+  max_tokens?: number;
+
+  /**
+   * Presence penalty parameter
+   */
+  presence_penalty?: number;
+
+  /**
+   * Repetition penalty parameter
+   */
+  repetition_penalty?: number;
+
+  /**
+   * Stop sequences
+   */
+  stop?: Array<string>;
+
+  /**
+   * Whether to stream the response
+   */
+  stream?: boolean;
+
+  /**
+   * Options that control streaming behavior.
+   */
+  stream_options?: CompletionCreateParams.StreamOptions;
+
+  /**
+   * The suffix to complete
+   */
+  suffix?: string;
+
+  /**
+   * Top-k sampling parameter. `-1` disables the cutoff.
+   */
+  top_k?: number;
+
+  /**
+   * Float that controls the cumulative probability of the top tokens to consider
+   */
+  top_p?: number;
+}
+
+export namespace CompletionCreateParams {
+  /**
+   * Options that control streaming behavior.
+   */
+  export interface StreamOptions {
+    /**
+     * If true, an additional chunk is streamed before the `data: [DONE]` message
+     * containing the token usage statistics for the entire request. Inside that chunk,
+     * `choices` is an empty array and the `usage` field is populated.
+     */
+    include_usage?: boolean;
+  }
+}
+
+export declare namespace Completions {
+  export { type CompletionCreateParams as CompletionCreateParams };
+}