--- title: "Get all partial chats for a chat session" method: GET path: "/chat/{chat_session_id}/partial-chats" tags: ["chat"] --- # Get all partial chats for a chat session `GET /chat/{chat_session_id}/partial-chats` ## Path parameters - `chat_session_id` string, required ## Response `200` Partial chats retrieved successfully - ChatPrompt[] - `decoding_params` DecodingParams, required - `parameters` DecodingParamEnum[], required - union - object - `MaxTokens` integer, required - object - `MaxCompletionTokens` integer, required - object - `TopP` number, float, required - object - `RepeatWindow` integer, required - object - `RepeatPenalty` number, float, required - object - `Temperature` number, float, required - object - `StopTokens` string[], required - object - `LogitBias` object, required - object - `Functions` object[], required - object - `JsonValidator` object, required - object - `RegexValidator` string, required - object - `ContextFreeGrammar` string, required - object - `NumBeams` integer, required - object - `Crop` boolean, required - object - `Thinking` integer, required — Thinking tokens for Claude 3.7. Contains the budget in tokens for thinking. - object - `Verbosity` 'low' | 'medium' | 'high', required - object - `ReasoningEffort` 'low' | 'medium' | 'high' | 'minimal', required - `messages` Message[], required - `content` MessageContent[], required — We want this to be a vec of contents so we can accurately capture an interleaving of images and text. This is meant to be a completely raw, unprocessed representation of the text. Don't take stuff out. - union - object - `Text` string, required - object - `Image` string, binary, required - `role` 'user' | 'system' | 'assistant', required - `metadata` PromptMeta, required — All metadata required to generate a prompt for the LLM - `dataset_descriptor` DatasetDescriptor, required — A dataset is where you put multiple referential schemas. A dataset is a complete namespace where all references between schemas are held within the dataset. - `description` string, required - `llm_override_field` string, nullable - `name` string, required - `relationships` RelationshipDescriptor[], required - `description` string, required - `merge_strategy` RelationshipMergeStrategy - `source_cardinality_given_target_match` integer, nullable — Describes the expected cardinality of the source table when a match is found in the target table For example, if we have a source company and a target funding round, we expect the source company to appear in multiple funding rounds, but not *too* many. So if we have a funding round match, the expected number of unique companies is relatively small. This is an estimate of that number. - `target_cardinality_given_source_match` integer, nullable — Describes the expected cardinality of the target table when a match is found in the source table For example, if we have a source company and a target funding round, we usually expect some number of funding rounds to be associated with a single company but not *too* many. So if we have a company match, the expected number of unique funding rounds is relatively small. This is an estimate of that number. - `name` string, required - `properties` PropertyDescriptor[] - `description` string, required - `merge_strategy` union - 'Unique' — Property with unique 1:1 correspondence to its parent. Merge based on this property 100% of the time - object - … - 'NoSignal' - `name` string, required - `prop_type` union - 'String' - 'Boolean' - object - … - 'Integer' - 'Float' - 'Date' - 'URL' - 'Money' - 'Image' - 'PersonName' - 'Address' - `source_table` string, required - `target_table` string, required - `tables` TableDescriptor[], required - `description` string, required - `expected_cardinality` integer, nullable — Expected number of unique values in the complete dataset. This is used for our probabilistic merge strategy. - `name` string, required — Organized in a name, description format. - `primary_column` string, nullable - `properties` PropertyDescriptor[], required — Organized in a name, description format. - `description` string, required - `merge_strategy` union - 'Unique' — Property with unique 1:1 correspondence to its parent. Merge based on this property 100% of the time - object - … - 'NoSignal' - `name` string, required - `prop_type` union - 'String' - 'Boolean' - object - … - 'Integer' - 'Float' - 'Date' - 'URL' - 'Money' - 'Image' - 'PersonName' - 'Address' - `extracted_entities` LLMKnowledgeGraph[], required - `entities` LLMEntity[], required - `id` integer, required - `properties` object, required - `type` string, required - `relationships` LLMRelationship[] - `properties` object - `source` integer, required - `target` integer, required - `type` string, required - `extraction_criteria` SaveRequirement[], required - union — It's an OR statement across these. - RequiredRelationship - `relationship_name` string, required - RequiredEntity - `entity_id` string, uuid - `seeded_entity_id` integer, required — The integer id corresponding to an entity in the seeded entity graph (different from the global dataset entity id) - RequiredProperty - `property_names` string[], required — If there are multiple properties, it can match just one of them - `table_name` string, required — The table name of the entity to update - `formatter_specific` union, required - object - `ImageMeta` ImageMeta, required - `document_name` string, nullable - `document_page` integer, nullable - `image` string, nullable, required - `ocr_content` string, nullable - object - `WebMeta` WebMeta, required - `flags` Flag[], required - `ariaLabel` string, required - `height` number, double - `href` string, nullable - `isInteractive` boolean, nullable - `number` integer, nullable — The number by which the flag is referred in image, prompt, and tool calls. - `text` string - `type` string, required - `width` number, double — The serde default here is to give us backwards compatibility it's fine for these to be anything as long as the image isn't given since it won't regenerate. - `x` number, double, required - `y` number, double, required - `ocr_content` string, nullable - `screenshot` string, binary, nullable - `url` string, required - object - `TextMeta` TextMeta, required - `text` string, required - object - `ScraperMeta` ScraperMeta, required - `html_content` string, required - `url` string, required - `qa_potentially_sus_response` string, nullable - `tool_metadata` ToolMetadata[], required - `description` string, required - `name` 'Exit' | 'Save' | 'Wait' | 'Type' | 'Scroll' | 'ScrollToBottom' | 'Click' | 'Hover' | 'Error' | 'Google', required - `regex_validator` string, required ## Other responses - `403` — Access denied - `404` — Chat session not found --- [API](https://skmtc.net/structifyai/apis/structify.md) · [All operations](https://skmtc.net/structifyai/apis/structify/llms.txt) · [OpenAPI document](https://skmtc-service-staging.skmtc.workers.dev/v1/apis/structifyai/structify/versions/1364930b6980/schema)