Skip to main content
POST
Python (SDK)

Authorizations

tql_api_key
string
header
required

Headers

Connect-Protocol-Version
enum<number>
default:1
required

Define the version of the Connect protocol

Available options:
1
Connect-Timeout-Ms
number

Define the timeout, in ms

Body

application/json
paradigm
paradigm · object

ChatParadigm includes paradigm options

model
enum<string>
Available options:
MODEL_UNKNOWN,
MODEL_DEFAULT_SMALL,
MODEL_DEFAULT,
MODEL_DEFAULT_LARGE,
MODEL_DEFAULT_REASONING,
MODEL_HAIKU_4_5,
MODEL_OPUS_4_8,
MODEL_FABLE_5,
MODEL_SONNET_5,
MODEL_GPT_5_6_SOL,
MODEL_GEMINI_3_FLASH,
MODEL_GEMINI_3_PRO,
MODEL_GEMINI_3_1_PRO,
MODEL_GEMINI_3_5_FLASH,
MODEL_GPT_5_6_TERRA,
MODEL_GPT_5_6_LUNA,
MODEL_DEEPSEEK_3_2,
MODEL_GLM_5,
MODEL_VLLM,
MODEL_KIMI_K2_6,
MODEL_GLM_5_2,
MODEL_KIMI_K2_7_CODE,
MODEL_QWEN3_7_PLUS,
MODEL_MUSE_SPARK_1_1
message
string | null

optionally pre-fill first message

playbookId
string | null

optionally associate with a playbook

research
boolean | null

whether to enable report mode for this chat

dashboardMode
boolean | null

whether to enable dashboard mode for this chat

methodology
enum<string> | null

optionally set chat methodology at creation

Available options:
METHODOLOGY_UNKNOWN,
METHODOLOGY_ADAPTIVE,
METHODOLOGY_PRESCRIPTIVE,
METHODOLOGY_THOROUGH,
METHODOLOGY_CAREFUL,
METHODOLOGY_ONTOLOGY_BUILDING
vllmModelId
string | null

vllm_model_id is the model identifier forwarded to the org's vLLM endpoint. Only valid when model == MODEL_VLLM. Requires @textql.com superadmin.

fastMode
boolean | null

fast_mode enables Anthropic's fast inference (speed: "fast") for this chat. Currently supported on Opus 4.6 only. Pricing is 6x standard rates.

maxThinking
boolean | null

max_thinking runs extended thinking at max effort with visible reasoning. Supported on Sonnet 5, Fable 5, and Opus 4.8.

Response

Success

chat
chat · object