Skip to main content
Add these nodes with the pipeline builder: pipeline.add(name="...").<node>(...). Each entry lists the node’s configuration parameters. See the Pipeline reference for add, run, and lifecycle methods.

​
ai_audio_operations

Leverage AI for Audio operations 
pipeline.add(name="node").ai_audio_operations()
Parameters
​
sub_type
str
default:"''"

​
ai_fill_pdf

Fill out a PDF with form fields using AI. The AI will understand and fill each field using provided context. To convert your PDF to have fillable input fields, use: https://www.sejda.com/pdf-forms 
pipeline.add(name="node").ai_fill_pdf(provider="anthropic", context="...", file=...)
Parameters
​
select_pages
bool
default:"False"
Whether to select specific pages to fill
​
provider
str
required
The model provider
​
model
str
default:"''"
The specific model for filling the PDF
​
context
str
required
Context used by LLM to fill PDF fields
​
file
AcceptsFile
required
The PDF with form fields to be filled
​
selected_pages
str
default:"''"
PDF page range

​
ai_image_operations

Leverage AI for Image operations 
pipeline.add(name="node").ai_image_operations()
Parameters
​
sub_type
str
default:"''"

​
ai_image_to_image

Modify and edit images using AI by providing modification instructions
Platform docs: Modify and edit images using AI by providing modification instructions
pipeline.add(name="node").ai_image_to_image(provider="google", model="gemini-2.5-flash-image", api_key="...", size="1024x1024")
Parameters
​
use_personal_api_key
bool
default:"False"
Use your personal API key
​
provider
str
required
Select the model provider. One of: google, openai
​
model
str
required
Select the image-to-image model
​
prompt
str
default:"''"
Tell the AI model how you would like it to modify the images. Be as specific as possible. For example, you can instruct the model to change colors, add elements, apply artistic styles, or blend multiple images. Must not be empty.
​
api_key
str
required
Input your personal API key from the model provider. Note: if you do not have access to the selected model, the workflow will not run
​
aspect_ratio
str
default:"'1:1'"
Select the aspect ratio for the output image.
​
images
AcceptsImageList
Array of input images to modify. Provide 1-3 images for best results.
​
size
str
required
Select the size. One of: 1024x1024, 1024x1536, 1024x1792, 1536x1024, 1792x1024, 256x256, 512x512, auto

​
ai_image_to_text

Generate Text from Image using AI
Platform docs: Generate Text from Image using AI
pipeline.add(name="node").ai_image_to_text(provider="anthropic", model="chatgpt-4o-latest", api_key="...", image=...)
Parameters
​
use_personal_api_key
bool
default:"False"
Use your personal API key
​
json_response
bool
default:"False"
Return the response as a JSON object
​
stream
bool
default:"False"
Stream the response
​
provider
str
required
Select the provider that will be used to analyze image. One of: anthropic, fireworks, google, openai, together, xai
​
model
str
required
Select the image analyzing model.
​
prompt
str
default:"''"
Instructions on what you want to analyze from the image.
​
system
str
default:"''"
Tell the AI model how you would like it to respond. Be as specific as possible. For example, you can instruct the model on what tone to respond in or how to respond given the information you provide
​
api_key
str
required
Input your personal API key from the model provider. Note: if you do not have access to the selected model, the workflow will not run
​
image
AcceptsImage
required
The image to analyze. For agent tool calls, pass an existing file reference such as history.MESSAGEID,history.MESSAGE_ID, tool.CALL_ID.OUTPUT_KEY, or $input.NAME; never pass an empty string.
​
json_schema
str
required
The JSON schema to use for the response
​
sampling
SamplingConfig
​
max_tokens
int
The maximum number of tokens to generate
​
temperature
float
The temperature of the model
​
top_p
float
The top-p value

​
ai_operations

Leverage AI for various tasks 
pipeline.add(name="node").ai_operations()
Parameters
​
sub_type
str
default:"''"

​
ai_routing

Route the execution flow to different paths based on the value of the conditions.
Platform docs: Route the execution flow to different paths based on the value of the conditions.
pipeline.add(name="node").ai_routing(provider="anthropic", input_query="...")
Parameters
​
provider
str
required
The model provider
​
enable_shared_memory
bool
default:"False"
If enabled, Shared Memory will be included as context for routing decisions
​
model
str
default:"''"
The specific model for categorization
​
conditions
list[str]
default:"[]"
​
input_query
str
required
Specify the message that the AI router will classify
​
outputs
dict
default:"{}"
​
shared_memory
str
default:"''"
The shared memory that will be used as context for AI to determine the routing
​
sampling
SamplingConfig
​
max_tokens
int
The maximum number of tokens to generate
​
temperature
float
The temperature of the model
​
top_p
float
The top-p value

​
ai_speech_to_text

Generate Text from Audio using AI
Platform docs: Generate Text from Audio using AI
pipeline.add(name="node").ai_speech_to_text(provider="deepgram", model="base", api_key="...", audio=...)
Parameters
​
use_personal_api_key
bool
default:"False"
Use your personal API key
​
provider
str
required
Select the model provider. One of: deepgram, google, openai
​
model
str
required
Select the speech-to-text model
​
api_key
str
required
Input your personal API key from the model provider. Note: if you do not have access to the selected model, the workflow will not run
​
audio
AcceptsAudio
required
The audio for conversion
​
tier
str
required
Select the tier

​
ai_text_to_image

Generate Image from Text using AI
Platform docs: Generate Image from Text using AI
pipeline.add(name="node").ai_text_to_image(provider="flux", model="Stable Diffusion 1.6", api_key="...", aspect_ratio="16:9")
Parameters
​
use_personal_api_key
bool
default:"False"
Use your personal API key
​
provider
str
required
Select the model provider. One of: flux, google, openai, stabilityai, xai
​
model
str
required
Select the text-to-image model
​
prompt
str
default:"''"
Tell the AI model how you would like it to respond. Be as specific as possible. For example, you can instruct the model to use bright colors. Must not be empty.
​
api_key
str
required
Input your personal API key from the model provider. Note: if you do not have access to the selected model, the workflow will not run
​
aspect_ratio
str
required
Select the aspect ratio.
​
size
str
required
Select the size. One of: 1024x1024, 1024x1536, 1024x1792, 1536x1024, 1792x1024, 256x256, 512x512, auto

​
ai_text_to_speech

Generate Audio from text using AI
Platform docs: Generate Audio from text using AI
pipeline.add(name="node").ai_text_to_speech(text="...", api_key="...")
Parameters
​
use_personal_api_key
bool
default:"False"
Use your personal API key
​
provider
str
default:"'openai'"
Select the model provider. One of: eleven_labs, openai
​
model
str
default:"'tts-1-hd'"
Select the text-to-speech model
​
text
str
required
The string input for conversion.
​
api_key
str
required
Input your personal API key from the model provider. Note: if you do not have access to the selected model, the workflow will not run
​
voice
str
default:"'alloy'"
Select the voice

​
categorizer

Categorize text using AI into custom-defined buckets 
pipeline.add(name="node").categorizer(provider="anthropic", text="...")
Parameters
​
justification
bool
default:"False"
Include the AI’s justification for its score
​
provider
str
required
The model provider
​
model
str
default:"''"
The specific model for categorization
​
text
str
required
The text that will be categorized
​
additional_context
str
default:"''"
Provide any additional context or instructions
​
fields
ListType | list[List[Dict[str, Any]]] | list[NameDescription]
default:"[]"
The fields to be categorized
​
sampling
SamplingConfig
​
max_tokens
int
The maximum number of tokens to generate
​
temperature
float
The temperature of the model
​
top_p
float
The top-p value

​
email_validator

Validate an email address 
pipeline.add(name="node").email_validator(api_key="...", email_to_validate="...")
Parameters
​
model
str
default:"'regex'"
The validation model to use One of: custom-validator, regex
​
api_key
str
required
The API key to use
​
email_to_validate
str
required
The email you want to validate
​
provider
str
default:"'hunter'"
The validation provider to use One of: debounce, hunter

​
extract_data

Extract key pieces of information or a list of information from a input text. 
pipeline.add(name="node").extract_data(text="...", fields=..., processed_outputs=...)
Parameters
​
model
str
default:"'gpt-4o'"
​
text
str
required
​
additional_context
str
default:"''"
​
fields
ListType | list[List[Dict[str, Any]]] | list[NameTypeDescription]
required
​
processed_outputs
dict
required
​
provider
str
default:"'openai'"

​
image_gen

pipeline.add(name="node").image_gen()
Parameters
​
model
str
default:"'gpt-4-1106-preview'"
​
prompt
str
default:"''"
​
aspect_ratio
str
default:"'1:1'"
​
image_count
str
default:"'1'"
​
provider
str
default:"'llmOpenAI'"
​
size
str
default:"'512x512'"

​
llm

LLM
Platform docs: LLM
pipeline.add(name="node").llm(provider="anthropic", api_key="...", json_schema="...")
Parameters
​
provider
str
required
Select the LLM provider to be used
​
stream
bool
default:"False"
Whether to stream the response
​
use_personal_api_key
bool
default:"True"
​
json_response
bool
default:"False"
Whether to return the response as a JSON object
​
show_sources
bool
default:"True"
Whether to show the sources used to generate the response
​
model
str
default:"''"
Select the LLM model to be used
​
prompt
str
default:"''"
The data that is sent to the LLM. Add data from other nodes with double curly braces e.g., {{input_0.text}}
​
system
str
default:"''"
The system prompt to be used
​
api_key
str
required
Your API key
​
baseUrl
str
default:"''"
​
citation_metadata
list[str]
default:"[]"
The metadata of the sources used to generate the response
​
enable_web_search
bool
default:"False"
Enable Claude’s built-in web search tool to search the web during response generation
​
jsonSchema
str
default:"''"
​
json_schema
str
required
The schema of the JSON response
​
reasoning_effort
str
default:"'default'"
Controls the depth of reasoning for GPT-5 models (“none” is only supported on GPT-5.1 variants). One of: default, high, low, medium, minimal, none
​
safe_context_token_window
bool
default:"False"
If enabled, the context window will be reduced to fit the model’s maximum context window.
​
verbosity
str
default:"'default'"
Controls the verbosity of GPT-5 responses. One of: default, high, low, medium
​
sampling
SamplingConfig
​
safety
SafetyConfig
​
retry
RetryConfig
​
enable_moderation
bool
Whether to enable moderation
​
enable_pii_address
bool
Whether to enable PII address
​
enable_pii_cc
bool
Whether to enable PII cc
​
enable_pii_email
bool
Whether to enable PII email
​
enable_pii_name
bool
Whether to enable PII name
​
enable_pii_phone
bool
Whether to enable PII phone
​
enable_pii_ssn
bool
Whether to enable PII ssn
​
max_retries
int
The maximum number of retries
​
max_tokens
int
The maximum amount of input + output tokens the model will take in and generate per run (1 token = 4 characters). Note: different models have different token limits and the workflow will error if the max token is reached.
​
retry_interval_ms
int
The interval between retries in milliseconds
​
retry_on_failure
bool
Enable retrying when the node execution fails
​
temperature
float
The “creativity” of the response - increase the temperature for more creative responses.
​
top_p
float
The “randomness” of the output - higher Top P values increase the randomness
​
deployment_id
str
default:"''"
The deployment ID for the Azure OpenAI model. This is required when using Azure OpenAI services.
​
endpoint
str
default:"''"
The Azure OpenAI endpoint URL (e.g., https://your-resource-name.openai.azure.com)
​
thinking_token_limit
int
default:"24576"
The maximum number of tokens the model can use for thinking
​
finetuned_model
str
default:"''"
Use your finetuned model for response generation. Make sure to select the matching base model from the dropdown.
​
use_finetuned_model
bool
default:"False"
​
base_url
str
default:"''"
The base URL of the custom LLM provider
​
aws_access_key_id
str
default:"''"
Your AWS Access Key ID
​
aws_region
str
default:"'us-east-1'"
AWS region where Bedrock models are enabled
​
aws_secret_access_key
str
default:"''"
Your AWS Secret Access Key

​
llm_anthropic_vision

pipeline.add(name="node").llm_anthropic_vision(api_key="...", image=..., json_schema="...")
Parameters
​
use_personal_api_key
bool
default:"False"
​
json_response
bool
default:"False"
​
model
str
default:"'claude-3-haiku-20240307'"
​
prompt
str
default:"''"
​
system
str
default:"''"
​
api_key
str
required
​
image
AcceptsImage
required
​
json_schema
str
required
​
sampling
SamplingConfig
​
max_tokens
int
​
temperature
float
​
top_p
float

​
llm_google_vision

pipeline.add(name="node").llm_google_vision(prompt="...", api_key="...", image=...)
Parameters
​
model
str
default:"'gemini-pro-vision'"
​
prompt
str
required
​
api_key
str
required
​
image
AcceptsImage
required
​
json_response
bool
default:"False"
​
provider
str
default:"'googleImageToText'"
​
stream
bool
default:"False"
​
sampling
SamplingConfig

​
llm_open_ai_vision

pipeline.add(name="node").llm_open_ai_vision(prompt="...", system="...", api_key="...", image=...)
Parameters
​
provider
str
default:"'openAiImageToText'"
​
stream
bool
default:"False"
​
use_personal_api_key
bool
default:"False"
​
json_response
bool
default:"False"
​
model
str
default:"'gpt-4-vision-preview'"
​
prompt
str
required
​
system
str
required
​
api_key
str
required
​
image
AcceptsImage
required
​
json_schema
str
required
​
sampling
SamplingConfig
​
max_tokens
int
​
temperature
float
​
top_p
float

​
nl_to_sql

Convert natural language queries to SQL queries. 
pipeline.add(name="node").nl_to_sql(text="...", db_dialect="MySQL", schema="...")
Parameters
​
model
str
default:"'gpt-4o'"
​
text
str
required
​
db_dialect
str
required
One of: MySQL, PostgreSQL, SQLite3, Snowflake
​
schema
str
required

​
reducto_extract

Extract structured data from documents using Reducto 
pipeline.add(name="node").reducto_extract(json_schema="...", system_prompt="...")
Parameters
​
use_personal_api_key
bool
default:"False"
Use your own Reducto API key instead of the platform default.
​
return_citations
bool
default:"False"
Return citation bounding boxes for extracted fields.
​
api_key
str
default:"''"
Your personal Reducto API key.
​
deep_extract
bool
default:"True"
Enable agentic deep extraction for higher accuracy. Uses iterative verification against the source material.
​
files
AcceptsFileList
default:"[]"
Documents to extract data from (up to 2,500 pages per document).
​
json_schema
str
required
A JSON schema defining the structure of data to extract. Use descriptive field names.
​
system_prompt
str
required
Instructions for how the AI should extract and verify data from the documents.

​
sales_data_enrichment

Enhance sales data with validation and enrichment processes 
pipeline.add(name="node").sales_data_enrichment()
Parameters
​
sub_type
str
default:"''"

​
scorer

Assign a numerical score based on predefined criteria to quantitatively assess a given entity. 
pipeline.add(name="node").scorer(provider="anthropic", text="...", criteria="...")
Parameters
​
justification
bool
default:"False"
Include the AI’s justification for its score
​
provider
str
required
The model provider
​
model
str
default:"''"
The specific model for scoring
​
text
str
required
The text that will be scored
​
additional_context
str
default:"''"
Provide any additional context or instructions
​
criteria
str
required
The criteria that the text will be scored
​
sampling
SamplingConfig
​
max_tokens
Any
​
temperature
Any
​
top_p
Any

​
speech_to_text

pipeline.add(name="node").speech_to_text(audio=...)
Parameters
​
model
str
default:"'OpenAI Whisper'"
One of: Deepgram, OpenAI Whisper
​
audio
AcceptsAudio
required
​
submodel
str
default:"'nova-2'"
​
tier
str
default:"'general'"

​
summarizer

Summarize large texts using AI. 
pipeline.add(name="node").summarizer(provider="anthropic", text="...")
Parameters
​
provider
str
required
The model provider
​
model
str
default:"''"
The specific model for summarization
​
text
str
required
The text to be summarized
​
sampling
SamplingConfig
​
max_tokens
Any
​
temperature
Any
​
top_p
Any

​
translator

Translate text from one language to another. 
pipeline.add(name="node").translator(provider="anthropic", text="...")
Parameters
​
provider
str
required
The model provider
​
model
str
default:"''"
The specific model for translation
​
text
str
required
The text to be translated
​
source_language
str
default:"'Detect Language'"
The language of the input text
​
target_language
str
default:"'English'"
The language to translate to

​
tts_eleven_labs

pipeline.add(name="node").tts_eleven_labs(text="...", api_key="...")
Parameters
​
model
str
default:"'eleven_multilingual_v2'"
​
text
str
required
​
api_key
str
required
​
voice
str
default:"'shimmer'"

​
tts_open_ai

pipeline.add(name="node").tts_open_ai(text="...", api_key="...")
Parameters
​
use_personal_api_key
bool
default:"False"
​
model
str
default:"'tts-1-hd'"
​
text
str
required
​
api_key
str
required
​
voice
str
default:"'alloy'"