Integration

Leap integration

Leap runs image, video and audio generation models (Veo, Kling, Seedance, FLUX, GPT Image, ElevenLabs and more) and ready-made presets behind one request shape, paid from one prepaid balance. Browse models and read each one's input schema, price a run before starting it, start a generation (the tool waits for the result) or a batch of up to 50, keep an output as a file, manage kept characters, webhook endpoints and events, and read your balance. The keyless public model routes (same catalog as list_models), the default and rotated webhook signing secret routes, and the three-step video or sound upload flow (a presigned PUT) are not exposed. Authentication is a workspace API key (leap_...) sent in the x-api-key header.

Connect Leap How integrations work

Ask it in chat

Say what should happen.
The workflow builds itself.

When this happens, do that in Leap.Built from the 29 Leap tools below, on the connection you add once.

What you can do with Leap

Each of these is one operation on one endpoint, available both as a workflow step and as a tool an AI agent can call on its own. Nothing bundles several operations into one action, so what a run did stays readable afterwards.

Most endpoints run on the free plan. A few, mainly the ones that publish on your behalf, need a paid plan.

Character, 5 operations

create_characterCreate characterPOSTKeeps a character from 1 to 4 pictures you uploaded (`upload_file`) or made (`keep_generation_output`). Leap reads its look from them once. Needs the generations:write scope.
delete_characterDelete characterDELETEDeletes a character. Its pictures stay as files in your workspace, and runs that named it keep their results. Needs an API key with the `generations:write` scope.
list_charactersList charactersGETYour workspace's characters, newest first, the ones made in the studio included. Needs an API key with the `generations:read` scope.
retrieve_characterRetrieve characterGETOne character, with fresh links to its pictures. Needs an API key with the `generations:read` scope.
update_characterUpdate characterPATCHRenames a character, or gives it new pictures in place of all the old ones; Leap then reads its look again. Runs that named it stay as they ran. Needs the generations:write scope.

Webhook endpoint, 5 operations

create_webhook_endpointCreate webhook endpointPOSTCreates a webhook endpoint that receives the workspace's events (every type by default). The url must be https and answer 2xx within 15 seconds. The signing secret is not returned here, so deliveries cannot be verified.
delete_webhook_endpointDelete webhook endpointDELETERemoves an endpoint. Its events stop going there. Needs an API key with the `generations:write` scope.
list_webhook_endpointsList webhook endpointsGETYour workspace's webhook endpoints, newest first. Their secrets are never listed. Needs an API key with the `generations:read` scope.
retrieve_webhook_endpointRetrieve webhook endpointGETOne endpoint, without its secret. Needs an API key with the `generations:read` scope.
update_webhook_endpointUpdate webhook endpointPATCHChanges any of `url`, `events` and `description` (null clears it), and turns deliveries off and on with `disabled`. Needs the generations:write scope.

File, 4 operations

delete_fileDelete fileDELETEDeletes a file you uploaded, and its bytes. Runs that already used it keep their results, and their `input` still names the deleted ID. Needs the generations:write scope.
keep_generation_outputKeep generation outputPOSTTurns an image output of a succeeded generation into a file of your workspace, so you can pass its ID to the next run without downloading and uploading it. Needs the generations:write scope.
retrieve_fileRetrieve fileGETA file's details, with a fresh signed link valid for 24 hours. Needs an API key with the `generations:read` scope.
upload_fileUpload filePOSTUploads one image (PNG, JPEG or WebP, up to 4 MB) and returns its file id (file_...) to pass in a model's input or to create_character. Videos and sounds cannot be uploaded through this integration. Needs the generations:write scope.

Generation, 3 operations

create_generationCreate generationPOSTRuns a model or preset: build input from retrieve_model's input_schema (create_quote prices it). Billed from the balance only if it succeeds. Waits for the job and returns the finished result; gives up after 10 minutes.
list_generationsList generationsGETYour workspace's generations, newest first, 50 to a page by default. Filter by model, status or where they were started. Needs the generations:read scope.
retrieve_generationRetrieve generationGETA generation's status, outputs and cost. Each read signs fresh output links, valid for 24 hours. Needs the generations:read scope.

Batch, 2 operations

create_batchCreate batchPOSTStarts up to 50 generations at once (all start or none do) and returns the batch with each generation queued. Read progress with retrieve_batch or retrieve_generation. Each request is a model and its input. Needs the generations:write scope.
retrieve_batchRetrieve batchGETReturns a batch with every generation in it (status, outputs, cost each). Send Prefer wait=60 to hold the read until every run is final or 60 seconds pass. Needs the generations:read scope.

Event, 2 operations

list_eventsList eventsGETYour workspace's events from the last 30 days, newest first, optionally only some types. Use it to catch up after your endpoint was down. Needs the generations:read scope.
retrieve_eventRetrieve eventGETOne event, with its `data.object` as the API returns it now, so its output links are fresh. Needs an API key with the `generations:read` scope.

Model, 2 operations

list_modelsList modelsGETEvery model and preset your key can run, with its price, in one answer. Filter it on your side by `modality` or `tasks`. Read one model for the JSON Schema of its input. Needs the generations:read scope.
retrieve_modelRetrieve modelGETReturns one model or preset with input_schema, the JSON Schema of the input object create_generation takes (required fields, allowed values, defaults), plus its pricing. Needs the generations:read scope.

Cancel, 1 operation

cancel_generationCancel generationPOSTStops a queued or running generation and releases its hold; nothing is charged. A generation that already ended comes back unchanged; a run that cannot be stopped once started answers 409. Needs the generations:write scope.

Credit, 1 operation

retrieve_creditsRetrieve creditsGETWhat your workspace can spend now, what running generations hold, whether auto top-up will add more when it runs low, and whether it got the welcome credit. Needs the generations:read scope.

Delivery, 1 operation

list_event_deliveriesList event deliveriesGETWhere the event went, newest first: one delivery per URL, each with every try, the status your server answered, and when Leap tries again. Needs the generations:read scope.

Quote, 1 operation

create_quoteCreate quotePOSTReturns the exact price (cost_usd, hold_usd) of a run without running it or holding credit. Takes the same model and input as create_generation and refuses the same invalid input (400). Needs the generations:read scope.

Redeliver, 1 operation

redeliver_eventRedeliver eventPOSTSends an event again, to one endpoint (endpoint we_...) or else to every enabled endpoint subscribed to its type and the webhook URL its request named. Returns the queued deliveries. Needs the generations:write scope.

Test, 1 operation

test_webhook_endpointTest webhook endpointPOSTSends a `webhook.test` event to this endpoint only, once, even if it's disabled, and answers with the delivery and how the try went. Needs the generations:write scope.

Use Leap from Claude Code, Cursor or Codex

Connect Leap in LiveContext, create an API key in Settings, then add the LiveContext MCP server to your client. Its Leap operations become tools the client can call, next to every other integration you connected.

Claude Code

claude mcp add --transport http livecontext https://livecontext.ai/mcp --header "X-API-Key: YOUR_API_KEY"

Codex (~/.codex/config.toml)

[mcp_servers.livecontext]
url = "https://livecontext.ai/mcp"

[mcp_servers.livecontext.http_headers]
"X-API-Key" = "YOUR_API_KEY"

Cursor, Claude Desktop and other JSON configs

{
  "mcpServers": {
    "livecontext": {
      "type": "http",
      "url": "https://livecontext.ai/mcp",
      "headers": {
        "X-API-Key": "YOUR_API_KEY"
      }
    }
  }
}

Set up the MCP server step by step

How to connect Leap

Create your workspace
Sign up free. Nothing to install.
Connect Leap once
Add your Leap connection in Settings. Every tool from this integration reuses it, and secrets are encrypted at rest and visible only to you.
Describe the job in chat
The workflow builds itself in front of you and picks the endpoints it needs. Drag the steps yourself on the canvas whenever you would rather be explicit.