API reference - Abstract

Item	Value
Base URL	`https://api.abstract.inc`
Authentication	`Authorization: Bearer` with your API key, on every endpoint
Request and answer bodies	JSON, server-sent events when `stream` is `true`, or a file’s bytes on the robot run file routes

These pages describe each request. They do not send requests. Run requests from your own machine, with the key in your program’s environment:

curl -s https://api.abstract.inc/v1/models -H "Authorization: Bearer $(abstract token)"

Every error has the same body. See Errors and limits.

Robot models

One observation in, one action chunk out.

Robot runs

Store, list, download and delete your robot runs.

LLM models

The OpenAI Responses API.

Robot models

Endpoint	Serves
`POST /v1/chat/completions`	One observation in, one action chunk out

The WebSocket at wss://api.abstract.inc/v1/responses runs a whole episode on one socket. OpenAPI does not describe it. See Sessions over WebSocket.

Robot runs

/v1/robot/runs stores your robot runs (episode logs, the Rerun recording with the camera video) under your account, and lists, downloads and deletes them. Only the key’s account reaches its runs. An account stores up to 100 GiB of runs.

Endpoint	Serves
`GET /v1/robot/runs`	List your completed runs
`GET /v1/robot/runs/usage`	What your runs hold against the quota, and this month’s metering
`GET /v1/robot/runs/{robot}/{run}`	A completed run’s manifest
`DELETE /v1/robot/runs/{robot}/{run}`	Delete every file of a run
`POST /v1/robot/runs/{robot}/{run}/complete`	Complete a run, so it is listed
`PUT /v1/robot/runs/{robot}/{run}/files/{path}`	Upload a file of at most 64 MiB
`GET /v1/robot/runs/{robot}/{run}/files/{path}`	Download a file
`DELETE /v1/robot/runs/{robot}/{run}/files/{path}`	Delete a file
`GET /v1/robot/runs/{robot}/{run}/uploads/{path}`	Resume a multipart upload
`POST /v1/robot/runs/{robot}/{run}/uploads/{path}`	Start or finish a multipart upload
`PUT /v1/robot/runs/{robot}/{run}/uploads/{path}`	Upload one part
`DELETE /v1/robot/runs/{robot}/{run}/uploads/{path}`	Abort a multipart upload

LLM models

Endpoint	Serves
`POST /v1/responses`	The OpenAI Responses API
`GET /v1/responses/{response_id}`	Retrieve a response your account created
`GET /v1/models`	The LLM model ids

Robot models

Robot runs

LLM models

​Robot models

​Robot runs

​LLM models

Robot models

Robot runs

LLM models