Changelog

Releases, fixes and API changes, published from the docs as they ship.

  1. each::workflows

    Optional workflow inputs can be left out

    Optional workflow inputs can be left out

    • An input marked "required": false no longer needs a default_value. Leave it out of the trigger and the steps that use it run without it.
    • A step parameter that refers to a left-out input is not sent to the model, and list items or object fields that refer to it are dropped.
    • In a choice condition a left-out input counts as missing; branch on it with not_exists.
    • A left-out input inside a text template still fails the step, and required inputs work as before.
    Read on docs
  2. API

    end_user_id is required on every model and workflow request

    end_user_id is required on every model and workflow request

    • Send end_user_id on prediction creation, workflow triggers, the OpenAI-compatible endpoints (chat, responses, audio) and Decisions. Do not send it to each::sense. These rules apply now.
    • It must identify the real person at the end of your chain of customers. If your customers are businesses, pass the ID of their user down to each::labs. It must be stable per person and pseudonymous, never a new random ID per request or session, and never a name, an email address, a phone number or an IP address, hashed or not.
    • Send it at the top level of the JSON body; on transcriptions it is a multipart form field. It must be 1 to 256 characters.
    • A refusal names its code: END_USER_BLOCKED or ORGANIZATION_BLOCKED (403), END_USER_ID_REQUIRED (400; a request without end_user_id can be refused at any time, as enforcement is switched on per organization without notice), SAFETY_BLOCKS_UNAVAILABLE (503, retryable), and SAFETY_PREFLIGHT_BLOCKED (422) or SAFETY_PREFLIGHT_UNAVAILABLE (503, retryable) from content safety checks.
    • Details are in End-user ID requirements, End-user ID reference and Safety Refusals.
    Read on docs
  3. LLM Router

    Run Decisions as a model and in workflows

    Run Decisions as a model

    • The Decisions API is now available as the eachlabs-decisions model: run it with POST /v1/prediction, from the playground, or as a workflow step that later steps branch on.
    • Runs through /v1/prediction are asynchronous, so you can poll the prediction or receive a webhook.
    • GET /v1/decisions/models lists the supported decision model IDs, and the playground's model dropdown uses the same list.
    Read on docs
  4. each::storage

    Delete files from the dashboard

    Delete files from the dashboard

    • Organization owners can delete a file from the Storage page (the ⋯ menu, then Delete) or from a run's page (the delete button under the output).
    • When a run produced several files, you choose which one to delete.
    • Deleting is permanent: the file's link stops working everywhere, including in runs and workflows that reused it.
    • Members who are not owners don't see delete controls. Deleting through the API with Delete File works as before.
    Read on docs
  5. MCP

    each::labs plugin for Claude and Codex

    each::labs plugin for Claude and Codex

    • The plugin bundles the each::labs MCP connector with twelve skills for common media tasks: image and video generation and editing, music, voice, fashion, product visuals and workflows.
    • Install it from the eachlabs/plugin marketplace: under Customize > Plugins in claude.ai and Claude Desktop, with claude plugin install eachlabs@eachlabs in Claude Code, and with codex plugin add eachlabs@eachlabs in Codex.
    • Sign-in, billing and the tools are the same as with the connector alone. Steps are on the Claude and Codex pages.
    Read on docs
  6. MCP

    each::labs MCP server: run models and workflows, and use interactive cards

    each::labs MCP server: run models and workflows, and use interactive cards

    • Run any each::labs model, not only video tools: search_models and get_model_schema find a model and its inputs, and run_model runs it, with up to 8 variants of one request.
    • Run your organization's workflows: list_workflows, get_workflow, trigger_workflow and get_workflow_execution.
    • Interactive cards in AI apps with a chat interface that support MCP Apps: upload files, view and download results, choose one of several variants, adjust a model's settings, fill in a workflow's inputs, compare an edit with the original, and trim or crop by dragging. Apps without a chat interface, such as terminals, have no cards: they show results as links and ask you in the chat instead.
    • A long job comes back as running with its id; check it with get_job, or get_workflow_execution for a workflow, until it is done.
    • New list_recent_jobs lists your organization's recent runs (id, model, status, time and result link) so the AI can pick up a run it started earlier.
    • New permissions workflows:run, workflows:read and uploads:create. Existing connections keep working for the video tools; reconnect to approve the new permissions and see the new tools.
    • Details are on the MCP Server page and in the MCP tool reference. Using it from Claude: Use each::labs in Claude.
    Read on docs
  7. LLM Router

    Use any OpenRouter chat model with the LLM Router

    Use any OpenRouter chat model with the LLM Router

    • POST /v1/chat/completions accepts any chat model ID from OpenRouter's model list. Send the ID exactly as listed.
    • Available Models now lists our featured models and every other available model, and shows the matching OpenRouter ID when a featured model's ID differs.
    • Pricing is the model's list price plus the same standard platform fee OpenRouter charges, nothing more.
    • The model catalog API reference moved to each::router.
    Read on docs
  8. Webhooks

    Redeliver a webhook from the dashboard

    Redeliver a webhook from the dashboard

    • Open a webhook on the Webhooks page and choose Redeliver to send the same payload to your endpoint again, without re-running the execution.
    • The button is offered only when a redelivery would be accepted, and names the reason when it would not: a delivery still in flight, a short cooldown after a previous redelivery, or a payload that is no longer stored.
    • A redelivery carries X-Webhook-Redelivery: true, which an ordinary delivery never sends, so your endpoint can tell the two apart.
    • Signed webhooks are signed again when the redelivery is sent, so it passes your timestamp tolerance rather than looking like a replayed request. See Redeliveries are signed again.
    • The body is unchanged, so handle deliveries idempotently and key on execution_id.
    Read on docs
  9. API

    Ask structured questions with the Decisions API

    Ask structured questions with the Decisions API

    • POST /v1/decisions answers typed questions over one shared context and returns structured results instead of prose: a noul question gives you a probability, a choice question a named option, and a score question a rating on an ordered scale.
    • Send model: "typesafe/jev-1.13" with state and questions, authenticated with the same Bearer API key as the rest of each::api. The response carries answers keyed by your question names plus billable usage.
    • The model is routed through OpenRouter to TypeSafe Jev. This API is in alpha and its contract may change — see the Decisions API page.
    Read on docs
  10. API

    Webhook deliveries are now signed with HMAC-SHA256

    Webhook deliveries are now signed with HMAC-SHA256

    • Set a webhook_secret when creating a prediction or triggering a workflow, and every delivery for that execution now carries X-Webhook-Signature and X-Webhook-Timestamp.
    • The signature is an HMAC-SHA256 over <X-Webhook-Timestamp>.<raw request body>, hex-encoded and prefixed with sha256=.
    • The new Verifying Webhook Signatures guide covers the procedure, with Python and JavaScript handlers.
    • The plain-text X-Webhook-Secret header is still sent, so receivers comparing it keep working. Verify the signature instead — it also proves the body was not modified.
    • Deliveries triggered without a webhook_secret are unchanged and remain unsigned.
    Read on docs
  11. API

    Model details now use a path-based endpoint

    Model details now use a path-based endpoint

    • GET /v1/models/{slug} is now the single route for retrieving model details. It returns the same payload as before, including the nullable cost guidance.
    • The deprecated GET /v1/model?slug= query route has been removed. Update any integration that used it to GET /v1/models/{slug}.
    Read on docs
  12. Organization Settings

    Control execution and storage retention for your organization

    Control execution and storage retention for your organization

    • Organization Settings now puts execution data and stored file retention in your hands. Choose how long each type of data is kept to match your team's needs.
    • Manage execution retention and storage retention separately, with automatic cleanup based on your organization's settings.
    • Review and update your retention preferences in one place as your team's data policies evolve.
    Read on docs
  13. LLM Router

    An easier LLM Router interface and errors by model

    An easier LLM Router interface and errors by model

    • LLM Router now uses the same model-aware form on model pages and in both workflow editors. Edit prompts, upload images, and adjust supported options through familiar controls that adapt to your selected model.
    • In Canvas Workflows, add and connect LLM Router like any other model card. The shared interface makes it easier to move from trying a model to using it in a workflow.
    • Error Analytics now breaks down LLM Router errors by the requested model, so you can see which models are failing and focus your investigation.
    Read on docs
  14. API

    Traverse execution history with cursor pagination

    Traverse execution history with cursor pagination

    • GET /v2/executions lists organization-wide execution history in descending creation order with the same filters as V1, a default limit of 20, and a maximum limit of 100. Authenticate with your API key as a Bearer token.
    • Follow the opaque next_cursor only while has_more is true. Keep the same filters and limit for each page; restart without a cursor to change them.
    • GET /v1/executions is deprecated but remains available for shallow compatibility. Requests beyond the 1,000-row prefix (offset <= 1000 - appliedLimit) return {"error":"V1 execution pagination is limited to 1000 rows; use GET /v2/executions with cursor pagination"}.
    Read on docs
  15. Billing

    Create spending alerts for your organization

    Create spending alerts for your organization

    • The dashboard now has Activity → Alerts for organization-wide spending alerts.
    • Create threshold rules for rolling charged spend, enable or disable them, and review the 20 most recent triggers from the same page.
    • Alert emails are sent to the organization owner and billing managers, with the charged spend, threshold, rolling window, and evaluation time included for reconciliation.
    Read on docs

Start with what you need.

Scale when you’re ready.