Package {aisdk.providers}

aisdk.providers: Additional Model Provider Adapters for the 'aisdk' Toolkit

Description

Additional AI model provider adapters for the 'aisdk' toolkit, covering OpenAI-compatible and Anthropic-compatible services such as 'DeepSeek', 'Moonshot'/'Kimi', 'Stepfun', 'Volcengine', 'AiHubMix', 'xAI', 'OpenRouter', 'Bailian', and 'NVIDIA'. Providers register themselves with the core 'aisdk' provider registry on load.

Author(s)

Maintainer: Yonghe Xia xiayh17@gmail.com

See Also

Useful links:

• https://github.com/YuLab-SMU/aisdk.providers

• Report bugs at https://github.com/YuLab-SMU/aisdk.providers/issues

AiHubMix Language Model Class

Description

Language model implementation for AiHubMix's chat completions API. Inherits from OpenAILanguageModel as AiHubMix provides an OpenAI-compatible API.

Super classes

aisdk::LanguageModelV1 -> aisdk::OpenAILanguageModel -> AiHubMixLanguageModel

Methods

Public methods

• AiHubMixLanguageModel$parse_response()

• AiHubMixLanguageModel$build_payload()

• AiHubMixLanguageModel$build_stream_payload()

• AiHubMixLanguageModel$clone()

Method parse_response()

Parse the API response into a GenerateResult. Overrides parent to extract AiHubMix-specific reasoning fields.

Usage

AiHubMixLanguageModel$parse_response(response)

Arguments

Returns

A GenerateResult object.

Method build_payload()

Build the request payload for non-streaming generation. Overrides parent to process caching and reasoning parameters.

Usage

AiHubMixLanguageModel$build_payload(params)

Arguments

Returns

A list with url, headers, and body.

Method build_stream_payload()

Build the request payload for streaming generation. Overrides parent to process caching and reasoning parameters.

Usage

AiHubMixLanguageModel$build_stream_payload(params)

Arguments

Returns

A list with url, headers, and body.

Method clone()

The objects of this class are cloneable with this method.

Usage

AiHubMixLanguageModel$clone(deep = FALSE)

Arguments

AiHubMix Provider Class

Description

Provider class for AiHubMix.

Super class

aisdk::OpenAIProvider -> AiHubMixProvider

Methods

Public methods

• AiHubMixProvider$new()

• AiHubMixProvider$language_model()

• AiHubMixProvider$image_model()

• AiHubMixProvider$clone()

Method new()

Initialize the AiHubMix provider.

Usage

AiHubMixProvider$new(
  api_key = NULL,
  base_url = NULL,
  headers = NULL,
  timeout_seconds = NULL,
  total_timeout_seconds = NULL,
  first_byte_timeout_seconds = NULL,
  connect_timeout_seconds = NULL,
  idle_timeout_seconds = NULL
)

Arguments

Method language_model()

Create a language model.

Usage

AiHubMixProvider$language_model(model_id = NULL)

Arguments

Returns

An AiHubMixLanguageModel object.

Method image_model()

Create an image model.

Usage

AiHubMixProvider$image_model(model_id = NULL)

Arguments

Returns

An OpenAIImageModel object.

Method clone()

The objects of this class are cloneable with this method.

Usage

AiHubMixProvider$clone(deep = FALSE)

Arguments

Bailian Language Model Class

Description

Language model implementation for Alibaba Cloud DashScope's chat completions API. Inherits from OpenAI model but adds support for DashScope-specific features like reasoning content extraction from Qwen reasoning models.

Super classes

aisdk::LanguageModelV1 -> aisdk::OpenAILanguageModel -> BailianLanguageModel

Methods

Public methods

• BailianLanguageModel$parse_response()

• BailianLanguageModel$clone()

Method parse_response()

Parse the API response into a GenerateResult. Overrides parent to extract DashScope-specific reasoning_content.

Usage

BailianLanguageModel$parse_response(response)

Arguments

Returns

A GenerateResult object.

Method clone()

The objects of this class are cloneable with this method.

Usage

BailianLanguageModel$clone(deep = FALSE)

Arguments

Bailian Provider Class

Description

Provider class for Alibaba Cloud Bailian / DashScope platform.

Super class

aisdk::OpenAIProvider -> BailianProvider

Methods

Public methods

• BailianProvider$new()

• BailianProvider$language_model()

• BailianProvider$clone()

Method new()

Initialize the Bailian provider.

Usage

BailianProvider$new(
  api_key = NULL,
  base_url = NULL,
  headers = NULL,
  timeout_seconds = NULL,
  total_timeout_seconds = NULL,
  first_byte_timeout_seconds = NULL,
  connect_timeout_seconds = NULL,
  idle_timeout_seconds = NULL
)

Arguments

Method language_model()

Create a language model.

Usage

BailianProvider$language_model(model_id = NULL)

Arguments

Returns

A BailianLanguageModel object.

Method clone()

The objects of this class are cloneable with this method.

Usage

BailianProvider$clone(deep = FALSE)

Arguments

DeepSeek Language Model Class

Description

Language model implementation for DeepSeek's chat completions API. Inherits from OpenAI model but adds support for DeepSeek-specific features like reasoning content extraction and DeepSeek thinking-mode parameters.

Super classes

aisdk::LanguageModelV1 -> aisdk::OpenAILanguageModel -> DeepSeekLanguageModel

Methods

Public methods

• DeepSeekLanguageModel$new()

• DeepSeekLanguageModel$parse_response()

• DeepSeekLanguageModel$build_payload()

• DeepSeekLanguageModel$build_stream_payload()

• DeepSeekLanguageModel$clone()

Method new()

Initialize the DeepSeek language model.

Usage

DeepSeekLanguageModel$new(model_id, config)

Arguments

Method parse_response()

Parse the API response into a GenerateResult. Overrides parent to extract DeepSeek-specific reasoning_content.

Usage

DeepSeekLanguageModel$parse_response(response)

Arguments

Returns

A GenerateResult object.

Method build_payload()

Build request payload with DeepSeek-specific reasoning params.

Usage

DeepSeekLanguageModel$build_payload(params)

Arguments

Returns

A list with url, headers, and body.

Method build_stream_payload()

Build stream payload with DeepSeek-specific reasoning params.

Usage

DeepSeekLanguageModel$build_stream_payload(params)

Arguments

Returns

A list with url, headers, and body.

Method clone()

The objects of this class are cloneable with this method.

Usage

DeepSeekLanguageModel$clone(deep = FALSE)

Arguments

DeepSeek Provider Class

Description

Provider class for DeepSeek.

Super class

aisdk::OpenAIProvider -> DeepSeekProvider

Methods

Public methods

• DeepSeekProvider$new()

• DeepSeekProvider$language_model()

• DeepSeekProvider$clone()

Method new()

Initialize the DeepSeek provider.

Usage

DeepSeekProvider$new(
  api_key = NULL,
  base_url = NULL,
  headers = NULL,
  timeout_seconds = NULL,
  total_timeout_seconds = NULL,
  first_byte_timeout_seconds = NULL,
  connect_timeout_seconds = NULL,
  idle_timeout_seconds = NULL
)

Arguments

Method language_model()

Create a language model.

Usage

DeepSeekProvider$language_model(model_id = NULL)

Arguments

Returns

A DeepSeekLanguageModel object.

Method clone()

The objects of this class are cloneable with this method.

Usage

DeepSeekProvider$clone(deep = FALSE)

Arguments

Kimi Code Anthropic Provider Class

Description

Anthropic-compatible provider wrapper for Kimi Code.

Super class

aisdk::AnthropicProvider -> KimiCodeAnthropicProvider

Methods

Public methods

• KimiCodeAnthropicProvider$language_model()

• KimiCodeAnthropicProvider$clone()

Method language_model()

Create a Kimi Code Anthropic-compatible language model.

Usage

KimiCodeAnthropicProvider$language_model(model_id = NULL)

Arguments

Returns

An AnthropicLanguageModel object.

Method clone()

The objects of this class are cloneable with this method.

Usage

KimiCodeAnthropicProvider$clone(deep = FALSE)

Arguments

Moonshot Language Model Class

Description

Language model implementation for Moonshot / Kimi Chat Completions APIs.

Super classes

aisdk::LanguageModelV1 -> aisdk::OpenAILanguageModel -> MoonshotLanguageModel

Methods

Public methods

• MoonshotLanguageModel$new()

• MoonshotLanguageModel$build_payload()

• MoonshotLanguageModel$build_stream_payload()

• MoonshotLanguageModel$parse_response()

• MoonshotLanguageModel$clone()

Method new()

Initialize the Moonshot language model.

Usage

MoonshotLanguageModel$new(model_id, config)

Arguments

Method build_payload()

Build the request payload for non-streaming generation.

Usage

MoonshotLanguageModel$build_payload(params)

Arguments

Returns

A list with url, headers, and body.

Method build_stream_payload()

Build the request payload for streaming generation.

Usage

MoonshotLanguageModel$build_stream_payload(params)

Arguments

Returns

A list with url, headers, and body.

Method parse_response()

Parse the API response into a GenerateResult.

Usage

MoonshotLanguageModel$parse_response(response)

Arguments

Returns

A GenerateResult object.

Method clone()

The objects of this class are cloneable with this method.

Usage

MoonshotLanguageModel$clone(deep = FALSE)

Arguments

Moonshot / Kimi Provider Class

Description

Provider class for Moonshot AI Kimi models.

Super class

aisdk::OpenAIProvider -> MoonshotProvider

Methods

Public methods

• MoonshotProvider$new()

• MoonshotProvider$language_model()

• MoonshotProvider$clone()

Method new()

Initialize the Moonshot provider.

Usage

MoonshotProvider$new(
  api_key = NULL,
  base_url = NULL,
  platform = c("auto", "platform", "coding"),
  headers = NULL,
  prompt_cache_key = NULL,
  timeout_seconds = NULL,
  total_timeout_seconds = NULL,
  first_byte_timeout_seconds = NULL,
  connect_timeout_seconds = NULL,
  idle_timeout_seconds = NULL
)

Arguments

Method language_model()

Create a language model.

Usage

MoonshotProvider$language_model(model_id = NULL)

Arguments

Returns

A MoonshotLanguageModel object.

Method clone()

The objects of this class are cloneable with this method.

Usage

MoonshotProvider$clone(deep = FALSE)

Arguments

NVIDIA Language Model Class

Description

Language model implementation for NVIDIA's chat completions API. Inherits from OpenAI model but adds support for NVIDIA-specific features like "enable_thinking" and reasoning content extraction.

Super classes

aisdk::LanguageModelV1 -> aisdk::OpenAILanguageModel -> NvidiaLanguageModel

Methods

Public methods

• NvidiaLanguageModel$parse_response()

• NvidiaLanguageModel$clone()

Method parse_response()

Parse the API response into a GenerateResult. Overrides parent to extract NVIDIA-specific reasoning_content.

Usage

NvidiaLanguageModel$parse_response(response)

Arguments

Returns

A GenerateResult object.

Method clone()

The objects of this class are cloneable with this method.

Usage

NvidiaLanguageModel$clone(deep = FALSE)

Arguments

NVIDIA Provider Class

Description

Provider class for NVIDIA.

Super class

aisdk::OpenAIProvider -> NvidiaProvider

Methods

Public methods

• NvidiaProvider$new()

• NvidiaProvider$language_model()

• NvidiaProvider$clone()

Method new()

Initialize the NVIDIA provider.

Usage

NvidiaProvider$new(
  api_key = NULL,
  base_url = NULL,
  headers = NULL,
  timeout_seconds = NULL,
  total_timeout_seconds = NULL,
  first_byte_timeout_seconds = NULL,
  connect_timeout_seconds = NULL,
  idle_timeout_seconds = NULL
)

Arguments

Method language_model()

Create a language model.

Usage

NvidiaProvider$language_model(model_id = NULL)

Arguments

Returns

A NvidiaLanguageModel object.

Method clone()

The objects of this class are cloneable with this method.

Usage

NvidiaProvider$clone(deep = FALSE)

Arguments

OpenRouter Language Model Class

Description

Language model implementation for OpenRouter's chat completions API. Inherits from OpenAI model but adds support for OpenRouter-specific features like reasoning content extraction from reasoning models.

Super classes

aisdk::LanguageModelV1 -> aisdk::OpenAILanguageModel -> OpenRouterLanguageModel

Methods

Public methods

• OpenRouterLanguageModel$parse_response()

• OpenRouterLanguageModel$clone()

Method parse_response()

Parse the API response into a GenerateResult. Overrides parent to extract reasoning_content from reasoning models.

Usage

OpenRouterLanguageModel$parse_response(response)

Arguments

Returns

A GenerateResult object.

Method clone()

The objects of this class are cloneable with this method.

Usage

OpenRouterLanguageModel$clone(deep = FALSE)

Arguments

OpenRouter Provider Class

Description

Provider class for OpenRouter.

Super class

aisdk::OpenAIProvider -> OpenRouterProvider

Methods

Public methods

• OpenRouterProvider$new()

• OpenRouterProvider$language_model()

• OpenRouterProvider$image_model()

• OpenRouterProvider$clone()

Method new()

Initialize the OpenRouter provider.

Usage

OpenRouterProvider$new(
  api_key = NULL,
  base_url = NULL,
  headers = NULL,
  timeout_seconds = NULL,
  total_timeout_seconds = NULL,
  first_byte_timeout_seconds = NULL,
  connect_timeout_seconds = NULL,
  idle_timeout_seconds = NULL
)

Arguments

Method language_model()

Create a language model.

Usage

OpenRouterProvider$language_model(model_id = NULL)

Arguments

Returns

An OpenRouterLanguageModel object.

Method image_model()

Create an image model.

Usage

OpenRouterProvider$image_model(model_id = NULL)

Arguments

Returns

An OpenAIImageModel object.

Method clone()

The objects of this class are cloneable with this method.

Usage

OpenRouterProvider$clone(deep = FALSE)

Arguments

Stepfun Image Model Class

Description

Image model implementation for Stepfun image generation and editing APIs.

Super class

aisdk::ImageModelV1 -> StepfunImageModel

Methods

Public methods

• StepfunImageModel$new()

• StepfunImageModel$do_generate_image()

• StepfunImageModel$do_edit_image()

• StepfunImageModel$clone()

Method new()

Initialize the Stepfun image model.

Usage

StepfunImageModel$new(model_id, config)

Arguments

Method do_generate_image()

Generate images.

Usage

StepfunImageModel$do_generate_image(params)

Arguments

Returns

A GenerateImageResult object.

Method do_edit_image()

Edit images.

Usage

StepfunImageModel$do_edit_image(params)

Arguments

Returns

A GenerateImageResult object.

Method clone()

The objects of this class are cloneable with this method.

Usage

StepfunImageModel$clone(deep = FALSE)

Arguments

Stepfun Language Model Class

Description

Language model implementation for Stepfun's chat completions API. Inherits from OpenAILanguageModel as Stepfun provides an OpenAI-compatible API.

Super classes

aisdk::LanguageModelV1 -> aisdk::OpenAILanguageModel -> StepfunLanguageModel

Methods

Public methods

• StepfunLanguageModel$build_payload()

• StepfunLanguageModel$build_stream_payload()

• StepfunLanguageModel$clone()

Method build_payload()

Build the payload for the Stepfun API.

Usage

StepfunLanguageModel$build_payload(params)

Arguments

Method build_stream_payload()

Build the stream payload for the Stepfun API.

Usage

StepfunLanguageModel$build_stream_payload(params)

Arguments

Method clone()

The objects of this class are cloneable with this method.

Usage

StepfunLanguageModel$clone(deep = FALSE)

Arguments

Stepfun Provider Class

Description

Provider class for Stepfun.

Super class

aisdk::OpenAIProvider -> StepfunProvider

Methods

Public methods

• StepfunProvider$new()

• StepfunProvider$language_model()

• StepfunProvider$image_model()

• StepfunProvider$clone()

Method new()

Initialize the Stepfun provider.

Usage

StepfunProvider$new(
  api_key = NULL,
  base_url = NULL,
  headers = NULL,
  timeout_seconds = NULL,
  total_timeout_seconds = NULL,
  first_byte_timeout_seconds = NULL,
  connect_timeout_seconds = NULL,
  idle_timeout_seconds = NULL
)

Arguments

Method language_model()

Create a language model.

Usage

StepfunProvider$language_model(model_id = NULL)

Arguments

Returns

A StepfunLanguageModel object.

Method image_model()

Create an image model.

Usage

StepfunProvider$image_model(model_id = NULL)

Arguments

Returns

A StepfunImageModel object.

Method clone()

The objects of this class are cloneable with this method.

Usage

StepfunProvider$clone(deep = FALSE)

Arguments

Volcengine Image Model Class

Description

Image model implementation for Volcengine Ark image generation models such as Doubao Seedream. Volcengine exposes these models through an OpenAI-compatible image generation endpoint.

Super class

aisdk::ImageModelV1 -> VolcengineImageModel

Methods

Public methods

• VolcengineImageModel$new()

• VolcengineImageModel$do_generate_image()

• VolcengineImageModel$do_edit_image()

• VolcengineImageModel$clone()

Method new()

Initialize the Volcengine image model.

Usage

VolcengineImageModel$new(model_id, config)

Arguments

Method do_generate_image()

Generate images.

Usage

VolcengineImageModel$do_generate_image(params)

Arguments

Returns

A GenerateImageResult object.

Method do_edit_image()

Edit images by providing one or more reference images.

Usage

VolcengineImageModel$do_edit_image(params)

Arguments

Returns

A GenerateImageResult object.

Method clone()

The objects of this class are cloneable with this method.

Usage

VolcengineImageModel$clone(deep = FALSE)

Arguments

Volcengine Language Model Class

Description

Language model implementation for Volcengine's chat completions API. Inherits from OpenAI model but adds support for Volcengine-specific features like reasoning content extraction from models that support reasoning_content.

Super classes

aisdk::LanguageModelV1 -> aisdk::OpenAILanguageModel -> VolcengineLanguageModel

Methods

Public methods

• VolcengineLanguageModel$parse_response()

• VolcengineLanguageModel$clone()

Method parse_response()

Parse the API response into a GenerateResult. Overrides parent to extract Volcengine-specific reasoning_content.

Usage

VolcengineLanguageModel$parse_response(response)

Arguments

Returns

A GenerateResult object.

Method clone()

The objects of this class are cloneable with this method.

Usage

VolcengineLanguageModel$clone(deep = FALSE)

Arguments

Volcengine Provider Class

Description

Provider class for the Volcengine Ark platform.

Super class

aisdk::OpenAIProvider -> VolcengineProvider

Methods

Public methods

• VolcengineProvider$new()

• VolcengineProvider$language_model()

• VolcengineProvider$image_model()

• VolcengineProvider$clone()

Method new()

Initialize the Volcengine provider.

Usage

VolcengineProvider$new(
  api_key = NULL,
  base_url = NULL,
  headers = NULL,
  timeout_seconds = NULL,
  total_timeout_seconds = NULL,
  first_byte_timeout_seconds = NULL,
  connect_timeout_seconds = NULL,
  idle_timeout_seconds = NULL
)

Arguments

Method language_model()

Create a language model.

Usage

VolcengineProvider$language_model(model_id = NULL)

Arguments

Returns

A VolcengineLanguageModel object.

Method image_model()

Create an image model.

Usage

VolcengineProvider$image_model(
  model_id = Sys.getenv("ARK_IMAGE_MODEL", "doubao-seedream-5-0")
)

Arguments

Returns

A VolcengineImageModel object.

Method clone()

The objects of this class are cloneable with this method.

Usage

VolcengineProvider$clone(deep = FALSE)

Arguments

xAI Image Model Class

Description

Image model implementation for xAI image generation and editing APIs.

Super class

aisdk::ImageModelV1 -> XAIImageModel

Methods

Public methods

• XAIImageModel$new()

• XAIImageModel$do_generate_image()

• XAIImageModel$do_edit_image()

• XAIImageModel$clone()

Method new()

Initialize the xAI image model.

Usage

XAIImageModel$new(model_id, config)

Arguments

Method do_generate_image()

Generate images.

Usage

XAIImageModel$do_generate_image(params)

Arguments

Returns

A GenerateImageResult object.

Method do_edit_image()

Edit images.

Usage

XAIImageModel$do_edit_image(params)

Arguments

Returns

A GenerateImageResult object.

Method clone()

The objects of this class are cloneable with this method.

Usage

XAIImageModel$clone(deep = FALSE)

Arguments

xAI Language Model Class

Description

Language model implementation for xAI's chat completions API. Inherits from OpenAILanguageModel as xAI provides An OpenAI-compatible API.

Super classes

aisdk::LanguageModelV1 -> aisdk::OpenAILanguageModel -> XAILanguageModel

Methods

Public methods

• XAILanguageModel$parse_response()

• XAILanguageModel$clone()

Method parse_response()

Parse the API response into a GenerateResult. Overrides parent to extract xAI-specific reasoning_content.

Usage

XAILanguageModel$parse_response(response)

Arguments

Returns

A GenerateResult object.

Method clone()

The objects of this class are cloneable with this method.

Usage

XAILanguageModel$clone(deep = FALSE)

Arguments

xAI Provider Class

Description

Provider class for xAI.

Super class

aisdk::OpenAIProvider -> XAIProvider

Methods

Public methods

• XAIProvider$new()

• XAIProvider$language_model()

• XAIProvider$image_model()

• XAIProvider$clone()

Method new()

Initialize the xAI provider.

Usage

XAIProvider$new(
  api_key = NULL,
  base_url = NULL,
  headers = NULL,
  timeout_seconds = NULL,
  total_timeout_seconds = NULL,
  first_byte_timeout_seconds = NULL,
  connect_timeout_seconds = NULL,
  idle_timeout_seconds = NULL
)

Arguments

Method language_model()

Create a language model.

Usage

XAIProvider$language_model(model_id = NULL)

Arguments

Returns

A XAILanguageModel object.

Method image_model()

Create an image model.

Usage

XAIProvider$image_model(model_id = NULL)

Arguments

Returns

A XAIImageModel object.

Method clone()

The objects of this class are cloneable with this method.

Usage

XAIProvider$clone(deep = FALSE)

Arguments

Create AiHubMix Provider

Description

Factory function to create an AiHubMix provider.

AiHubMix provides a unified API for various models including Claude, OpenAI, Gemini, etc.

Usage

create_aihubmix(
  api_key = NULL,
  base_url = NULL,
  headers = NULL,
  timeout_seconds = NULL,
  total_timeout_seconds = NULL,
  first_byte_timeout_seconds = NULL,
  connect_timeout_seconds = NULL,
  idle_timeout_seconds = NULL
)

Arguments

Value

An AiHubMixProvider object.

Examples

if (interactive()) {
    aihubmix <- create_aihubmix()
    model <- aihubmix$language_model("claude-sonnet-3-5")
    result <- generate_text(model, "Explain quantum computing in one sentence.")
}

Create AiHubMix Provider (Anthropic API Format)

Description

Factory function to create an AiHubMix provider using the Anthropic-compatible API. This allows you to use AiHubMix Claude models with the native Anthropic API format, unlocking advanced features like Prompt Caching.

Usage

create_aihubmix_anthropic(
  api_key = NULL,
  extended_caching = FALSE,
  headers = NULL,
  timeout_seconds = NULL,
  total_timeout_seconds = NULL,
  first_byte_timeout_seconds = NULL,
  connect_timeout_seconds = NULL,
  idle_timeout_seconds = NULL
)

Arguments

Details

AiHubMix provides an Anthropic-compatible endpoint at ⁠https://aihubmix.com/v1⁠. This convenience function wraps create_anthropic() with AiHubMix-specific defaults.

Value

An AnthropicProvider object configured for AiHubMix.

Examples

if (interactive()) {
    # Use AiHubMix via Anthropic API format (unlocks caching)
    aihubmix_claude <- create_aihubmix_anthropic()
    model <- aihubmix_claude$language_model("claude-3-5-sonnet-20241022")
    result <- generate_text(model, "Hello Claude!")
}

Create AiHubMix Provider (Gemini API Format)

Description

Factory function to create an AiHubMix provider using the Gemini-compatible API. This allows you to use Gemini models with the native Gemini API structure.

Usage

create_aihubmix_gemini(
  api_key = NULL,
  headers = NULL,
  timeout_seconds = NULL,
  total_timeout_seconds = NULL,
  first_byte_timeout_seconds = NULL,
  connect_timeout_seconds = NULL,
  idle_timeout_seconds = NULL
)

Arguments

Details

AiHubMix provides a Gemini-compatible endpoint at ⁠https://aihubmix.com/gemini/v1beta/models⁠. This convenience function wraps create_gemini() with AiHubMix-specific defaults.

Value

A GeminiProvider object configured for AiHubMix.

Examples

if (interactive()) {
    # Use AiHubMix via Gemini API format
    aihubmix_gemini <- create_aihubmix_gemini()
    model <- aihubmix_gemini$language_model("gemini-2.5-flash")
    result <- generate_text(model, "Hello Gemini!")
}

Create Alibaba Cloud Bailian Provider

Description

Factory function to create an Alibaba Cloud Bailian provider using the DashScope API.

Usage

create_bailian(
  api_key = NULL,
  base_url = NULL,
  headers = NULL,
  timeout_seconds = NULL,
  total_timeout_seconds = NULL,
  first_byte_timeout_seconds = NULL,
  connect_timeout_seconds = NULL,
  idle_timeout_seconds = NULL
)

Arguments

Value

A BailianProvider object.

Supported Models

• qwen3.6-plus: Latest Qwen flagship with 1M context. Strong reasoning and multimodal capabilities. (Reasoning, Vision, Tools, Structured) | ctx: 1000k

• qwen3.5-plus: Balanced model with excellent reasoning at ultra-low cost. (Reasoning, Vision, Tools, Structured) | ctx: 1000k

• qwen-plus: Most popular general-purpose model. Good balance of capability and cost. (Reasoning, Vision, Tools, Structured) | ctx: 1000k

• qwen-max: Previous generation premium model for complex reasoning and analysis. (Reasoning, Vision, Tools, Structured) | ctx: 128k

• qwen-turbo: Fast and cost-efficient model for high-volume simple tasks. (Vision, Tools, Structured) | ctx: 128k

• qwen-coder-plus: Coding specialist optimized for agentic software engineering. (Reasoning, Tools, Structured) | ctx: 128k

• qwq-32b: 32B parameter reasoning model with strong math and logic capabilities. (Reasoning, Tools, Structured) | ctx: 128k

• qwen-vl-plus: Vision-language model for image understanding and visual reasoning. (Vision, Tools, Structured) | ctx: 32k

• qwen3.5-omni-plus: Multimodal model supporting text, image, video, and audio input. (Reasoning, Vision, Tools, Audio, Structured) | ctx: 32k

• text-embedding-v3: Text embedding model for similarity search and retrieval. | ctx: 8k

Examples

if (interactive()) {
bailian <- create_bailian()

# Standard chat model
model <- bailian$language_model("qwen-plus")
result <- generate_text(model, "Hello")

# Reasoning model (QwQ with chain-of-thought)
model <- bailian$language_model("qwq-32b")
result <- generate_text(model, "Solve: What is 15 * 23?")
print(result$reasoning) # Chain-of-thought reasoning

# Default model (qwen-plus)
model <- bailian$language_model()
}

Create DeepSeek Provider

Description

Factory function to create a DeepSeek provider.

Usage

create_deepseek(
  api_key = NULL,
  base_url = NULL,
  headers = NULL,
  timeout_seconds = NULL,
  total_timeout_seconds = NULL,
  first_byte_timeout_seconds = NULL,
  connect_timeout_seconds = NULL,
  idle_timeout_seconds = NULL
)

Arguments

Details

DeepSeek supports classic aliases plus newer model families such as DeepSeek V4.

Common model IDs include:

• deepseek-chat: Chat alias provided by DeepSeek

• deepseek-reasoner: Reasoning alias provided by DeepSeek

• deepseek-v4*: DeepSeek V4 family model IDs exposed by the API

Additional DeepSeek-specific request fields such as thinking, thinking_budget, and reasoning_effort are passed through when supplied to ⁠$generate()⁠ or ⁠$stream()⁠.

Value

A DeepSeekProvider object.

Supported Models

• deepseek-v4-flash: DeepSeek V4 Flash hybrid-thinking model (Reasoning, Tools, Structured) | ctx: 1000k

• deepseek-v4-pro: DeepSeek V4 Pro hybrid-thinking model (Reasoning, Tools, Structured) | ctx: 1000k

Examples

if (interactive()) {
# Basic usage with deepseek-chat
deepseek <- create_deepseek()
model <- deepseek$language_model("deepseek-chat")
result <- generate_text(model, "Hello!")

# Using a reasoning-capable model
model_reasoner <- deepseek$language_model("deepseek-reasoner")
result <- model_reasoner$generate(
    messages = list(list(role = "user", content = "Solve: What is 15 * 23?")),
    max_tokens = 500,
    thinking = TRUE
)
print(result$text) # Final answer
print(result$reasoning) # Chain-of-thought reasoning

# Streaming with reasoning
stream_text(model_reasoner, "Explain quantum entanglement step by step")
}

Create DeepSeek Provider (Anthropic API Format)

Description

Factory function to create a DeepSeek provider using the Anthropic-compatible API. This allows you to use DeepSeek models with the Anthropic API format.

Usage

create_deepseek_anthropic(
  api_key = NULL,
  headers = NULL,
  timeout_seconds = NULL,
  total_timeout_seconds = NULL,
  first_byte_timeout_seconds = NULL,
  connect_timeout_seconds = NULL,
  idle_timeout_seconds = NULL
)

Arguments

Details

DeepSeek provides an Anthropic-compatible endpoint at ⁠https://api.deepseek.com/anthropic⁠. This convenience function wraps create_anthropic() with DeepSeek-specific defaults.

Note: When using an unsupported model name, the API backend will automatically map it to deepseek-chat.

Value

An AnthropicProvider object configured for DeepSeek.

Examples

if (interactive()) {
# Use DeepSeek via Anthropic API format
deepseek <- create_deepseek_anthropic()
model <- deepseek$language_model("deepseek-chat")
result <- generate_text(model, "Hello!")

# This is useful for tools that expect Anthropic API format
# such as Claude Code integration
}

Create Kimi Code Provider

Description

Convenience wrapper for Kimi Code membership API. By default this uses the Anthropic-compatible endpoint because it works for self-built coding agents with their real User-Agent. Set api_format = "openai" when integrating with OpenAI-compatible tools that Kimi Code recognizes as coding agents.

Usage

create_kimi_code(
  api_key = NULL,
  base_url = NULL,
  api_format = c("anthropic", "openai"),
  headers = NULL,
  prompt_cache_key = NULL,
  timeout_seconds = NULL,
  total_timeout_seconds = NULL,
  first_byte_timeout_seconds = NULL,
  connect_timeout_seconds = NULL,
  idle_timeout_seconds = NULL
)

Arguments

Value

A provider object configured for Kimi Code.

Create Kimi Code Provider (Anthropic API Format)

Description

Convenience wrapper for Kimi Code's Anthropic-compatible endpoint. Use model ID ⁠kimi-for-coding⁠. The public Kimi docs list the Anthropic base as ⁠https://api.kimi.com/coding/⁠; aisdk's Anthropic provider appends ⁠/messages⁠ directly, so this wrapper normalizes to ⁠https://api.kimi.com/coding/v1⁠.

Usage

create_kimi_code_anthropic(
  api_key = NULL,
  base_url = NULL,
  headers = NULL,
  timeout_seconds = NULL,
  total_timeout_seconds = NULL,
  first_byte_timeout_seconds = NULL,
  connect_timeout_seconds = NULL,
  idle_timeout_seconds = NULL
)

Arguments

Value

An AnthropicProvider object configured for Kimi Code.

Create Moonshot / Kimi Provider

Description

Factory function to create a Moonshot provider. Use platform = "platform" for the pay-as-you-go Kimi Open Platform (⁠https://api.moonshot.cn/v1⁠) and platform = "coding" for Kimi Code membership API (⁠https://api.kimi.com/coding/v1⁠). The two platforms use separate API keys.

Usage

create_moonshot(
  api_key = NULL,
  base_url = NULL,
  platform = c("auto", "platform", "coding"),
  headers = NULL,
  prompt_cache_key = NULL,
  timeout_seconds = NULL,
  total_timeout_seconds = NULL,
  first_byte_timeout_seconds = NULL,
  connect_timeout_seconds = NULL,
  idle_timeout_seconds = NULL
)

Arguments

Value

A MoonshotProvider object.

Supported Models

• kimi-k2.6: Kimi flagship multimodal model for agentic coding, long-context reasoning, and general agent tasks (Reasoning, Vision, Tools, Structured) | ctx: 262k

• kimi-k2.5: Kimi K2.5 multimodal model for agent, coding, vision, thinking, and conversation tasks (Reasoning, Vision, Tools, Structured) | ctx: 262k

• kimi-k2-0905-preview: Kimi K2 preview model with enhanced agentic coding, frontend, and context understanding capabilities (Reasoning, Tools, Structured) | ctx: 262k

• kimi-for-coding: Stable Kimi Code membership model ID for coding agents; backend maps it to the latest coding model (Reasoning, Vision, Tools, Structured) | ctx: 262k

• moonshot-v1-8k: Moonshot V1 text generation model with 8k context (Tools, Structured) | ctx: 8k

• moonshot-v1-32k: Moonshot V1 text generation model with 32k context (Tools, Structured) | ctx: 33k

• moonshot-v1-128k: Moonshot V1 text generation model with 128k context (Tools, Structured) | ctx: 131k

Examples

if (interactive()) {
moonshot <- create_moonshot()
model <- moonshot$language_model("kimi-k2.6")
result <- generate_text(model, "Hello", temperature = 1)

kimi_code <- create_moonshot(platform = "coding")
coding_model <- kimi_code$language_model()
result <- generate_text(coding_model, "Review this function", prompt_cache_key = "task-1")
}

Create NVIDIA Provider

Description

Factory function to create a NVIDIA provider.

Usage

create_nvidia(
  api_key = NULL,
  base_url = NULL,
  headers = NULL,
  timeout_seconds = NULL,
  total_timeout_seconds = NULL,
  first_byte_timeout_seconds = NULL,
  connect_timeout_seconds = NULL,
  idle_timeout_seconds = NULL
)

Arguments

Value

A NvidiaProvider object.

Examples

if (interactive()) {
nvidia <- create_nvidia()
model <- nvidia$language_model("z-ai/glm4.7")

# Enable thinking/reasoning
result <- generate_text(model, "Who are you?",
  chat_template_kwargs = list(enable_thinking = TRUE)
)
print(result$reasoning)
}

Create OpenRouter Provider

Description

Factory function to create an OpenRouter provider.

Usage

create_openrouter(
  api_key = NULL,
  base_url = NULL,
  headers = NULL,
  timeout_seconds = NULL,
  total_timeout_seconds = NULL,
  first_byte_timeout_seconds = NULL,
  connect_timeout_seconds = NULL,
  idle_timeout_seconds = NULL
)

Arguments

Value

An OpenRouterProvider object.

Supported Models

• openai/gpt-5.5: OpenAI's latest frontier model via OpenRouter. (Reasoning, Vision, Tools, Structured) | ctx: 1000k

• openai/gpt-5.4: OpenAI's previous flagship with computer use and 1M context. (Reasoning, Vision, Tools, Structured) | ctx: 1000k

• anthropic/claude-opus-4-7: Anthropic's most capable model via OpenRouter. (Reasoning, Vision, Tools, Structured) | ctx: 1000k

• anthropic/claude-sonnet-4-6: Balanced Claude model with 1M context and strong coding. (Reasoning, Vision, Tools, Structured) | ctx: 1000k

• google/gemini-3.1-pro: Google's most capable multimodal model via OpenRouter. (Reasoning, Vision, Tools, Audio, Structured, Search) | ctx: 1000k

• google/gemini-2.5-flash: Fast and cost-efficient Gemini with 1M context. (Reasoning, Vision, Tools, Audio, Structured, Search) | ctx: 1000k

• xai/grok-4.1-fast: xAI's fastest model with 2M context at aggressive pricing. (Reasoning, Vision, Tools, Structured, Search) | ctx: 2000k

• xai/grok-4: xAI frontier model with always-on reasoning. (Reasoning, Vision, Tools, Structured, Search) | ctx: 256k

• deepseek/deepseek-chat: DeepSeek's general-purpose chat model. (Vision, Tools, Structured) | ctx: 64k

• deepseek/deepseek-reasoner: DeepSeek's reasoning specialist (R1). (Reasoning, Tools, Structured) | ctx: 64k

• meta-llama/llama-4-maverick: Meta's open-weight frontier model with 10M context (Scout). (Reasoning, Vision, Tools, Structured) | ctx: 10000k

• qwen/qwen3.6-plus: Alibaba's latest Qwen flagship with 1M context. (Reasoning, Vision, Tools, Structured) | ctx: 1000k

• moonshotai/kimi-k2: Moonshot's agentic swarm model with 256K context. (Reasoning, Vision, Tools, Structured) | ctx: 256k

• openai/gpt-4.1: OpenAI's production workhorse with 1M context. (Vision, Tools, Structured) | ctx: 1000k

• anthropic/claude-haiku-4-5: Fast and cost-effective Claude for simple tasks. (Vision, Tools, Structured) | ctx: 200k

Examples

if (interactive()) {
openrouter <- create_openrouter()

# Access any model via a unified API
model <- openrouter$language_model("openai/gpt-4o")
result <- generate_text(model, "Hello!")

# Reasoning model
model <- openrouter$language_model("deepseek/deepseek-r1")
result <- generate_text(model, "Solve: 15 * 23")
print(result$reasoning)
}

Create Stepfun Provider

Description

Factory function to create a Stepfun provider.

Usage

create_stepfun(
  api_key = NULL,
  base_url = NULL,
  headers = NULL,
  timeout_seconds = NULL,
  total_timeout_seconds = NULL,
  first_byte_timeout_seconds = NULL,
  connect_timeout_seconds = NULL,
  idle_timeout_seconds = NULL
)

Arguments

Value

A StepfunProvider object.

Supported Models

• step-1-32k: Model: step-1-32k (Tools) | ctx: 32k

• step-1v-32k: Vision enabled model with 32k context (Vision, Tools) | ctx: 32k

• step-1-8k: Model: step-1-8k

• step-1-256k: Model: step-1-256k

• step-1v-8k: Model: step-1v-8k (Vision)

• step-2-16k: Model: step-2-16k

• step-1x-medium: Model: step-1x-medium

• step-tts-mini: Model: step-tts-mini (Audio)

• step-2-16k-202411: Model: step-2-16k-202411

• step-asr: Model: step-asr (Audio)

• step-1o-vision-32k: Model: step-1o-vision-32k (Vision)

• step-2-mini: Model: step-2-mini

• step-2-16k-exp: Model: step-2-16k-exp

• step-1o-turbo-vision: Model: step-1o-turbo-vision (Vision)

• step-1o-audio: Model: step-1o-audio (Audio)

• ... and 16 more models. Use list_models("stepfun") to see all.

Examples

if (interactive()) {
    stepfun <- create_stepfun()
    model <- stepfun$language_model("step-1-8k")
    result <- generate_text(model, "Explain quantum computing in one sentence.")
}

Create Volcengine/Ark Provider

Description

Factory function to create a Volcengine provider using the Ark API.

Usage

create_volcengine(
  api_key = NULL,
  base_url = NULL,
  headers = NULL,
  timeout_seconds = NULL,
  total_timeout_seconds = NULL,
  first_byte_timeout_seconds = NULL,
  connect_timeout_seconds = NULL,
  idle_timeout_seconds = NULL
)

Arguments

Value

A VolcengineProvider object.

Supported Models

• doubao-lite-128k-240428: Model: doubao-lite-128k-240428

• doubao-pro-128k-240515: Model: doubao-pro-128k-240515

• doubao-lite-4k-240328: Model: doubao-lite-4k-240328

• doubao-lite-32k-240428: Model: doubao-lite-32k-240428

• doubao-pro-4k-240515: Model: doubao-pro-4k-240515

• doubao-lite-4k-character-240515: Model: doubao-lite-4k-character-240515

• doubao-embedding-text-240515: Model: doubao-embedding-text-240515

• mistral-7b-instruct-v0.2: Model: mistral-7b-instruct-v0.2

• doubao-pro-4k-character-240515: Model: doubao-pro-4k-character-240515

• doubao-pro-4k-functioncall-240515: Model: doubao-pro-4k-functioncall-240515

• doubao-lite-4k-pretrain-character-240516: Model: doubao-lite-4k-pretrain-character-240516

• doubao-pro-32k-character-240528: Model: doubao-pro-32k-character-240528

• doubao-pro-4k-browsing-240524: Model: doubao-pro-4k-browsing-240524

• doubao-pro-32k-functioncall-240515: Model: doubao-pro-32k-functioncall-240515

• doubao-pro-4k-functioncall-240615: Doubao Pro function-calling model (4K); suited to low-latency function-calling. (Tools, Structured) | ctx: 4k

• ... and 95 more models. Use list_models("volcengine") to see all.

API Formats

Volcengine supports both Chat Completions API and Responses API:

• language_model(): Uses Chat Completions API (standard)

• responses_model(): Uses Responses API (for reasoning models)

• smart_model(): Auto-selects based on model ID

Token Limit Parameters for Volcengine Responses API

Volcengine's Responses API has two mutually exclusive token limit parameters:

• max_output_tokens: Total limit including reasoning + answer (default mapping)

• max_tokens (API level): Answer-only limit, excluding reasoning

The SDK's unified max_tokens parameter maps to max_output_tokens by default, which is the safe choice to prevent runaway reasoning costs.

For advanced users who want answer-only limits:

• Use max_answer_tokens parameter to explicitly set answer-only limit

• Use max_output_tokens parameter to explicitly set total limit

Examples

if (interactive()) {
    volcengine <- create_volcengine()

    # Chat API (standard models)
    model <- volcengine$language_model("doubao-1-5-pro-256k-250115")
    result <- generate_text(model, "Hello")

    # Responses API (reasoning models like DeepSeek)
    model <- volcengine$responses_model("deepseek-r1-250120")

    # Default: max_tokens limits total output (reasoning + answer)
    result <- model$generate(messages = msgs, max_tokens = 2000)

    # Advanced: limit only the answer part (reasoning can be longer)
    result <- model$generate(messages = msgs, max_answer_tokens = 500)

    # Smart model selection (auto-detects best API)
    model <- volcengine$smart_model("deepseek-r1-250120")
}

Create xAI Provider

Description

Factory function to create an xAI provider.

Usage

create_xai(
  api_key = NULL,
  base_url = NULL,
  headers = NULL,
  timeout_seconds = NULL,
  total_timeout_seconds = NULL,
  first_byte_timeout_seconds = NULL,
  connect_timeout_seconds = NULL,
  idle_timeout_seconds = NULL
)

Arguments

Value

A XAIProvider object.

Supported Models

• grok-4: xAI frontier model with always-on reasoning and multi-agent architecture. (Reasoning, Vision, Tools, Structured, Search) | ctx: 256k

• grok-4.1-fast: High-throughput production model with 2M context. Best default for most workloads. (Reasoning, Vision, Tools, Structured, Search) | ctx: 2000k

• grok-3: Legacy enterprise-grade model optimized for data extraction and summarization. (Reasoning, Vision, Tools, Structured, Search) | ctx: 131k

• grok-3-mini: Lightweight low-latency option with reasoning capabilities. (Reasoning, Tools, Structured, Search) | ctx: 131k

• grok-2-image: Image generation model via xAI API. (Image-Out) | ctx: 33k

Examples

if (interactive()) {
    xai <- create_xai()
    model <- xai$language_model("grok-beta")
    result <- generate_text(model, "Explain quantum computing in one sentence.")
}

AiHubMix Provider

Description

Implementation for AiHubMix models. AiHubMix API is OpenAI-compatible, but provides extended support for features like Claude's extended thinking and prompt caching.

Alibaba Cloud Bailian Provider

Description

Implementation for Alibaba Cloud Bailian (DashScope) hosted models. DashScope API is OpenAI-compatible with support for Qwen series models including reasoning models (QwQ, Qwen3 etc.).

DeepSeek Provider

Description

Implementation for DeepSeek models. DeepSeek API is OpenAI-compatible with support for reasoning models.

Moonshot / Kimi Provider

Description

Implementation for Moonshot AI Kimi models. The public Kimi Open Platform uses ⁠https://api.moonshot.cn/v1⁠; Kimi Code membership API uses ⁠https://api.kimi.com/coding/v1⁠. These are separate account systems and their API keys are not interchangeable.

NVIDIA Provider

Description

Implementation for NVIDIA NIM and other NVIDIA-hosted models.

OpenRouter Provider

Description

Implementation for OpenRouter, a unified API gateway for multiple LLM providers. OpenRouter API is OpenAI-compatible and provides access to models from OpenAI, Anthropic, Google, Meta, Mistral, DeepSeek, and many more.

Stepfun Provider

Description

Implementation for Stepfun models. Stepfun API is OpenAI-compatible.

Volcengine Provider

Description

Implementation for Volcengine Ark hosted models. Volcengine API is OpenAI-compatible with support for reasoning models (e.g., Doubao, DeepSeek).

xAI Provider

Description

Implementation for xAI (Grok) models. xAI API is OpenAI-compatible.