Lists the lab assays you can order, such as binding affinity or expression.

Each assay type lists its `settings`, in the same format `GET /tools` uses for a tool's
settings: a name, a type, whether it is required, and the `options` to choose from
when there is a fixed set. An order is for one assay type.

provider

Something went wrong on our side (`internal_error`). Try again.

Binding assays measure how well your sequences bind to a target. Use this to find a
target's id.

Put the id in your order's settings: `targetId` for Adaptyv and GenScript, `targetIds`
for A-Alpha Bio, or `targetAntigen` for Ginkgo. A GenScript target is one package of an
antigen, so the same antigen in two sizes is two targets.

Your account needs Assay Portal access to search targets.

The lab whose targets to search. An assay type's `targetsProvider` tells you which.

Text to look for in a target's id, name, aliases or UniProt id. Upper and lower case are treated the same.

search

limit

`provider` is missing or not one we offer, or `limit` is out of range

Your account doesn't have Assay Portal access yet (`wetlab_access_required`)

We couldn't load this lab's target list just now (`targets_unavailable`). Try again shortly.

Checks an order without placing it. Send the body you would send to
`POST /submit-assay`. Nothing is created, and your sequences are not sent to a lab.

The answer is a 200 whether or not the order is valid, so read `valid`. When it is
false, `error` says what is wrong, and `details` lists each problem when there are
several.

A valid order isn't compared with your earlier orders, so the real submit can still
come back with `duplicate_order`, `assays_blocked` or `assays_in_flight`.

Q3 binders round 2

binder-1

binder-2

The verdict. An order with problems is also a 200.

Your account doesn't have Assay Portal access yet (`wetlab_access_required`), or you are using a trial key that hasn't been claimed.

We couldn't load GenScript's antigen list to check the order (`targets_unavailable`). Nothing was created. Try again shortly.

Gives the estimated price and turnaround time for an order. Send the body you would
send to `POST /submit-assay`. No order is created and nothing is charged.

The price comes from the lab, so your sequences are sent to the lab to price them. It
is an estimate, not a quote. Tamarind confirms the
final price with you before your order is approved.

Some targets are priced on request. For those `estimatedCostUsd` is null and `note`
says so.

The order has a problem (`invalid_request`), or the lab would reject it or can't price it as it is (`quote_rejected`).

We couldn't get an estimate from the lab right now (`estimate_unavailable`). Try again shortly. You can place the order without one.

Places an order with a lab.

Tamarind reviews every order before a lab starts work on it. A new order starts as
`pending_approval`. We arrange payment with you, then approve the order. Nothing is
charged through the API.

Each order is for one assay type. To check an order without placing it, send the same
body to `POST /validate-assay`. For a price and turnaround estimate, send it to
`POST /estimate-assay`.

Send an `Idempotency-Key` header so that a retry can never place the same order twice.
If you repeat a request with the same key and body, you get the original order back.
A key is remembered for at least 7 days: retry within that window, and use a new key
for a new order after it.

When something goes wrong, the response includes a `code` your program can check, as
well as a message for people. The status codes below list the codes.

A value you choose that is unique to this order. Send the same one if you retry. Remembered for at least 7 days.

Idempotency-Key

Affinity

The order was placed. If you repeated a request with the same Idempotency-Key, this is the original order and `idempotentReplay` is true.

Something in the request is wrong. `error` says what, and `details` lists each problem when there are several.

Nothing was created. The `code` says why.

`idempotency_in_progress`: your earlier request with this key is still running. Wait
a minute and retry with the same key. A request that died can take about 6 minutes
to clear.

`idempotency_order_unavailable`: the order this key created no longer exists. Use a
new key to order again.

`duplicate_order`: you recently placed an order for the same assays and number of
sequences (see `duplicateOfOrderId`). If you want a second one, resend with
`confirmDuplicate: true`.

`assays_blocked`: an earlier attempt may already have reached the lab. We have been
notified and will follow up, so don't resubmit.

`assays_in_flight`: another submit for these assays is still running, or didn't
finish cleanly. Check GET /assays before trying again.

You already used this Idempotency-Key for a different order. Use a new key.

Something went wrong on our side. If the response has an `assayId`, the order was created, so don't resubmit. If it doesn't, check GET /assays; retrying with the same Idempotency-Key is safe. If you didn't send a key, check GET /assays before you retry, because a retry could place a second order.

A problem with the lab. The `code` says what happened.

`provider_error`: the lab failed after we recorded your order. We have been notified,
and the lab may be holding a draft.

`order_incomplete`: the order exists (see `assayId`) but didn't finish. We will
follow up, so don't resubmit.

`quote_unavailable`: we couldn't get a quote from A-Alpha, and nothing was created.
You can retry with the same key.

`targets_unavailable`: we couldn't load GenScript's antigen list to check the order,
and nothing was created. You can retry with the same key.

For `provider_error` and `order_incomplete`, retrying with the same key returns the
same answer.

A temporary problem, and nothing was created. Retry with the same key. The one exception is when `teamNotified` is true: then a retry won't help, and we will follow up.

Shows your assay orders, newest first. To look up one order, pass its `assayId`.

Orders come back a page at a time. When the response has a `nextToken`, send it back
to get the next page, and keep going until `nextToken` is null. A page can hold fewer
orders than `limit` and still have more to come.

An order normally goes from `pending_approval` to `approved`, then `in_progress`,
then `completed`.

Return just this order. The other parameters are ignored when you send it.

assayId

How many orders to return per page, up to 100.

The `nextToken` from the previous page, to get the next one.

nextToken

status

One order when you pass `assayId`. Otherwise a page of orders.

`limit`, `status` or `nextToken` isn't valid

You don't have an order with this `assayId`

Cancels an order.

If Tamarind hasn't approved the order yet, it is cancelled straight away. If it has
already gone to Ginkgo, it shows as `cancelling` until Ginkgo confirms. Adaptyv orders
can't be cancelled through the API, so contact Tamarind support for those.

Each order has a `cancellable` field that tells you whether it can be cancelled right
now. Cancelling an order that is already cancelled is fine: you just get it back.

The order, now `cancelled` or `cancelling`

`assayId` is missing, or the order can't be cancelled in its current state (`not_cancellable`). `status` shows where the order is.

The order can't be cancelled right now. The `code` says why.

`approval_in_progress`: Tamarind is approving it at this moment. Try again in a minute.

`approval_under_review`: an approval didn't finish cleanly and the team is checking
it. Contact support instead of retrying.

`order_changed`: the order changed while you were cancelling it. Look it up again.

We couldn't confirm the cancellation (`cancel_unconfirmed`). Check the order's status before trying again.

Gives you download links for an order's result files.

Labs deliver results some time after an order completes. Until they do, `available`
is false and `files` is empty, so check again later. Once results arrive they stay
available.

The links stop working after a while (`expiresInSeconds` is the longest they last), so
ask for fresh ones when you need them instead of saving them.

We couldn't load this order's results just now (`results_unavailable`). Try again shortly, or contact support if it keeps happening.

An assay order. You get one back when you submit, list or cancel orders.

Which assay the order is for. Null for some older orders placed in the web app.

The lab running the order, for example adaptyv, ginkgo, aalpha or genscript. More labs may be added, so don't assume these are the only values.

True once the lab has delivered results and GET /assay-results has files for you. It stays true from then on, even if the order's `status` changes later, so check this field instead of waiting for `status` to be `completed`.

An assay order. `POST /submit-assay`, `/validate-assay` and `/estimate-assay` all take this body.

Which assay to run. The options come from GET /assay-types.

Which CSV columns hold what. Needed with `csvFile`. For antibodies, map `name`, `heavy` and `light`. For proteins, map `name` and `sequence`.

Set to true to place an order that looks like one you placed recently. Use it after a `duplicate_order` response, when you do want a second, separately billed order.

A CSV file of sequences, to use instead of `sequences`. Upload it first with PUT /upload/{filename}, then give its path within your files.

Whether your sequences are antibodies or proteins. You can leave it out when the assay type only supports one.

A name for the order so you can find it later.

Anything the Tamarind team should know when reviewing the order.

The sequences to test. For antibodies, send "HEAVY:LIGHT" strings (or just "HEAVY"), or objects with `name`, `heavy` and `light`. For proteins, send the sequence as a string (join chains with ":"), or objects with `name` and `sequence`. A sequence you don't name is called seq_1, seq_2 and so on. Every name and every sequence has to be different.

True when `files` has every file the lab delivered.

Names of files the lab delivered that you can't download yet.

True when more targets matched than were returned. Narrow your search or raise `limit`.

Whether a protein can have several chains, joined with ":".

Set when this assay needs a target. Look targets up with GET /assay-targets, using this value as `provider`.

Order page

Single order

The estimated price divided by the number of sequences.

The usual turnaround for this assay type.

The lab's own turnaround estimate for this order in days

Present when not valid: `invalid_request`.

List custom tools in your organization.

Filter by status or publication state. Paginated — follow `nextCursor` until it is null.

Filter to published (`true`) or unpublished (`false`) tools.

published

Maximum number of tools to return in one page.

Pagination token from the previous response's `nextCursor`.

cursor

The request was malformed or failed validation; see `errors` for the offending fields.

An error occurred (RFC 9457 problem+json). Switch on the stable `code`, not on prose.

Create a custom tool in Draft status.

This creates the tool record only. Upload its source and build a version in separate requests.

Check a `config.json` against the contract a build applies, before spending a build on it.

Creates and changes nothing, and the tool need not exist. A config that fails is a successful
check: the response is 200 with `valid: false` and the reason a build would give. It does not
look at the rest of the source, so a build can still be refused for a missing `Dockerfile` or
`run.sh`.

Get a custom tool and one version.

By default this includes the latest version. Pass `version=v2` to inspect a specific version.
A tool that has never been built returns `version: null`.

name

version

The addressed resource does not exist, or is not visible to your tenant.

A supplied request condition did not match the current resource. Fetch the resource again before retrying.

Update a custom tool's metadata or compute resources.

Fields you omit are left unchanged.

Delete this custom tool and release its name for reuse.

Create a temporary upload for a source archive.

Upload the file to `uploadUrl` using `uploadMethod` and every value in `uploadHeaders`, then use
`uploadId` to build a version.

Build a version from an uploaded source archive. The archive also updates the tool’s editable source and configuration. Returns immediately with the new or reused version and its current status.

Retries the same build request without admitting duplicate work. Reusing a key with different source or runtime facts returns 409 Conflict.

The request conflicts with current resource state.

List build log events for a version.

By default this reads the latest version. Pass `version=v2` for a specific version. Paginated —
follow `nextCursor` until it is null.

Request cancellation of an active build. Omit version when exactly one build is active, or specify a version such as v2. Continue polling until the build reaches a terminal status.

Publish a completed version.

By default this publishes the latest completed version. Pass `version=v2` to publish or roll back
to a specific completed version.

The Tool with its published version updated.

List versions of a custom tool.

Filter by build status. Paginated — follow `nextCursor` until it is null.

Maximum number of versions to return in one page.

Code

Message

PublicBuildError

Timestamp

PublicBuildEvent

Items

Pass as `cursor` for the next page. Null when there are no more events.

Nextcursor

The version these log events belong to, such as `v3`.

Version

PublicBuildLogPage

What happened: `build` started a new image build, `reuse_image` created a version from the current image, and `unchanged` returned the version already associated with this source.

Action

PublicBuildResult

The first problem found, empty when valid. Fix it and validate again to find any others.

Errors

Whether a build would accept this `config.json`.

Valid

PublicConfigValidation

Number of CPU cores allocated to each run. Defaults to 1.

Short explanation of the tool. Defaults to an empty string.

Description

Human-readable name shown in the UI. Defaults to an empty string.

Displayname

GPU allocated to each run: "None", T4, L4, L40S, or A100. Defaults to "None". A10 is retired: it is still accepted and runs on L4.

Gputype

Memory allocated to each run. Defaults to "8Gi".

Memory

A unique lowercase name containing letters, numbers, and hyphens.

Name

PublicCreateCustomToolRequest

The `uploadId` returned when the source upload was created.

Uploadid

PublicCreateVersionRequest

Autopublish

Canbuild

Canedit

Createdat

The version used by default, such as `v3`. Null when no version is published.

Defaultversion

Esttime

Functions

Hassource

Homediskgi

Maximum runtime for a tool run, in seconds. Null means no tool-specific limit.

Maxruntimeseconds

Paperurl

Published

Updatedat

PublicCustomTool

The requested version, or the latest version when no version was requested.

PublicCustomToolDetail

Pass as `cursor` for the next page. Null when there are no more results.

PublicCustomToolPage

PublicCustomToolStatus

RFC 9457 problem detail. Serialized as `application/problem+json` on every public error.

A stable machine-readable slug; switch on THIS, not prose.

Detail

Structured per-item detail (request-validation fields OR pipeline diagnostics).

Whether the SAME request can succeed later. `true` means retry (honour `Retry-After` when present); `false` means this will not resolve on its own and retrying cannot change it. Omitted when the error does not specify — absence is NOT `false`.

Retryable

The HTTP status code, duplicated in the body per RFC 9457.

Status

A short, human-readable summary of the error kind.

Title

A URI identifying the error kind (dereferenceable docs).

Type

PublicProblem

Whether successful builds become the default version automatically. Omit or pass null to leave it unchanged.

Number of CPU cores allocated to each run. Omit or pass null to leave it unchanged.

Short explanation of the tool. Omit or pass null to leave it unchanged; pass an empty string to clear it.

Human-readable name shown in the UI. Omit or pass null to leave it unchanged; pass an empty string to clear it.

Estimated runtime in H:M:S format, for example "0:1:0". Minutes and seconds must be 0-59. Omit or pass null to leave it unchanged; pass an empty string to clear it.

Short capability labels used for discovery. Omit or pass null to leave them unchanged; pass an empty list to clear them.

GPU allocated to each run: "None", T4, L4, L40S, or A100. Omit or pass null to leave it unchanged. A10 is retired: it is still accepted and runs on L4.

Writable home-disk capacity in GiB. Omit or pass null to leave it unchanged.

Memory allocated to each run. Omit or pass null to leave it unchanged.

Related HTTP or HTTPS publication URL. Omit or pass null to leave it unchanged; pass an empty string to clear it.

PublicUpdateCustomToolRequest

When the upload URL expires, as an ISO 8601 timestamp.

Expiresat

Maxbytes

Headers to include exactly as provided when uploading to `uploadUrl`.

Uploadheaders

Pass this value to the version-build request after uploading.

Uploadmethod

Short-lived URL for uploading the source archive.

Uploadurl

PublicUploadSession

PublicValidateConfigRequest

Completedat

Origin

Immutable source revision used by this version. Null when source is hidden.

Sourcerevision

Startedat

Whether the build has reached a final status and will not run again.

Terminal

PublicVersion

PublicVersionPage

PublicVersionStatus

Recommend which model to run, given your inputs and natural language prompt, backed by Tamarind benchmarks. Currently supports structure prediction, more tasks coming soon.

Status code	Description
200	Successfully retrieved presigned URL to results
400	Bad request

API Documentation

Job Results

Required Parameters:

Optional Parameters:

HTTP Response Status Codes

Response Format