release: 0.40.0

.release-please-manifest.json

@@ -1,3 +1,3 @@
 {
-  ".": "0.39.0"
+  ".": "0.40.0"
 }

.stats.yml

@@ -1,4 +1,4 @@
 configured_endpoints: 75
-openapi_spec_url: https://storage.googleapis.com/stainless-sdk-openapi-specs/togetherai%2Ftogetherai-76f8801298719cc87e9dc4c64b321bcfd432416d76488499d340b4bb6bf81b9b.yml
-openapi_spec_hash: ce0b83ef0a5f174461bd7d13a379b636
+openapi_spec_url: https://storage.googleapis.com/stainless-sdk-openapi-specs/togetherai%2Ftogetherai-f5cce4e374b601983a3951d614fe8f964dc21b5ef913acdf6680e71393eb9af4.yml
+openapi_spec_hash: b91f36b9f8c40bc03b72ec5a73effcbc
 config_hash: 52d213100a0ca1a4b2cdcd2718936b51

CHANGELOG.md

@@ -1,5 +1,37 @@
 # Changelog
 
+## 0.40.0 (2026-04-23)
+
+Full Changelog: [v0.39.0...v0.40.0](https://github.com/togethercomputer/together-typescript/compare/v0.39.0...v0.40.0)
+
+### Features
+
+* **api:** accept string URLs for audio_inputs and source_video in videos ([9313762](https://github.com/togethercomputer/together-typescript/commit/93137622f3c3fe9158c5692a69a7a191aee27d45))
+* **api:** add cuda/nvidia driver fields, auto-scaling, OIDC, scheduled capacity to clusters ([881242f](https://github.com/togethercomputer/together-typescript/commit/881242f30f5436353c4c977d4a205df35aedb9d1))
+* **api:** add h100-40gb-mig and b200-192gb gpu_type options to jig resource ([dcadaf2](https://github.com/togethercomputer/together-typescript/commit/dcadaf20bdcd7a2d238a16e4ba7c807536e3dfab))
+* **api:** add max_seq_length parameter to fine-tuning ([2dd78f9](https://github.com/togethercomputer/together-typescript/commit/2dd78f9636d4a43d15a5fd1b180194d20684cca0))
+* **api:** add num_workers parameter to evals model/judge requests ([c61a7c1](https://github.com/togethercomputer/together-typescript/commit/c61a7c1cc7eafa0913d853e993992961cc38ebaf))
+
+
+### Bug Fixes
+
+* **types:** rename cuda_driver_version to cuda_version in clusters ([e83d0c4](https://github.com/togethercomputer/together-typescript/commit/e83d0c40631b352675cc6bc1149e86cfd9dbbcdd))
+* **types:** update driver_versions structure, require supported_instance_types in cluster regions ([d1caf1c](https://github.com/togethercomputer/together-typescript/commit/d1caf1cb6f030a67980ec7c5569fcf521745f191))
+
+
+### Chores
+
+* **internal:** codegen related update ([f9b079e](https://github.com/togethercomputer/together-typescript/commit/f9b079e57f54ed0f4e0d4671d298efef8543874d))
+* **internal:** codegen related update ([4ba1386](https://github.com/togethercomputer/together-typescript/commit/4ba1386fb9a20d7cd2a33c07c5e2fbb42e80d9fe))
+* **internal:** more robust bootstrap script ([5c001e2](https://github.com/togethercomputer/together-typescript/commit/5c001e23c5b364f21daf5d31e0111e9e15f40885))
+* **tests:** bump steady to v0.22.1 ([3d2bf61](https://github.com/togethercomputer/together-typescript/commit/3d2bf615258e63639612b29b0c544dc8c7c88b10))
+
+
+### Documentation
+
+* **api:** expand billing_type documentation in cluster creation ([f74418e](https://github.com/togethercomputer/together-typescript/commit/f74418e05cab6a9540d2482ffbecdec9fe0272c4))
+* improve examples ([bdf968a](https://github.com/togethercomputer/together-typescript/commit/bdf968a7f48622f0865040dd8f9b81d029ce8e3e))
+
 ## 0.39.0 (2026-04-03)
 
 Full Changelog: [v0.38.0...v0.39.0](https://github.com/togethercomputer/together-typescript/compare/v0.38.0...v0.39.0)

package.json

@@ -1,6 +1,6 @@
 {
   "name": "together-ai",
-  "version": "0.39.0",
+  "version": "0.40.0",
   "description": "The official TypeScript library for the Together API",
   "author": "Together <dev-feedback@TogetherAI.com>",
   "types": "dist/index.d.ts",

scripts/bootstrap

@@ -4,7 +4,7 @@ set -e
 
 cd "$(dirname "$0")/.."
 
-if [ -f "Brewfile" ] && [ "$(uname -s)" = "Darwin" ] && [ "$SKIP_BREW" != "1" ] && [ -t 0 ]; then
+if [ -f "Brewfile" ] && [ "$(uname -s)" = "Darwin" ] && [ "${SKIP_BREW:-}" != "1" ] && [ -t 0 ]; then
   brew bundle check >/dev/null 2>&1 || {
     echo -n "==> Install Homebrew dependencies? (y/N): "
     read -r response

scripts/mock

@@ -22,9 +22,9 @@ echo "==> Starting mock server with URL ${URL}"
 # Run steady mock on the given spec
 if [ "$1" == "--daemon" ]; then
   # Pre-install the package so the download doesn't eat into the startup timeout
-  npm exec --package=@stdy/cli@0.20.2 -- steady --version
+  npm exec --package=@stdy/cli@0.22.1 -- steady --version
 
-  npm exec --package=@stdy/cli@0.20.2 -- steady --host 127.0.0.1 -p 4010 --validator-query-array-format=comma --validator-form-array-format=comma --validator-query-object-format=brackets --validator-form-object-format=brackets "$URL" &> .stdy.log &
+  npm exec --package=@stdy/cli@0.22.1 -- steady --host 127.0.0.1 -p 4010 --validator-query-array-format=comma --validator-form-array-format=comma --validator-query-object-format=brackets --validator-form-object-format=brackets "$URL" &> .stdy.log &
 
   # Wait for server to come online via health endpoint (max 30s)
   echo -n "Waiting for server"
@@ -48,5 +48,5 @@ if [ "$1" == "--daemon" ]; then
 
   echo
 else
-  npm exec --package=@stdy/cli@0.20.2 -- steady --host 127.0.0.1 -p 4010 --validator-query-array-format=comma --validator-form-array-format=comma --validator-query-object-format=brackets --validator-form-object-format=brackets "$URL"
+  npm exec --package=@stdy/cli@0.22.1 -- steady --host 127.0.0.1 -p 4010 --validator-query-array-format=comma --validator-form-array-format=comma --validator-query-object-format=brackets --validator-form-object-format=brackets "$URL"
 fi

scripts/test

@@ -43,7 +43,7 @@ elif ! steady_is_running ; then
   echo -e "To run the server, pass in the path or url of your OpenAPI"
   echo -e "spec to the steady command:"
   echo
-  echo -e "  \$ ${YELLOW}npm exec --package=@stdy/cli@0.20.2 -- steady path/to/your.openapi.yml --host 127.0.0.1 -p 4010 --validator-query-array-format=comma --validator-form-array-format=comma --validator-query-object-format=brackets --validator-form-object-format=brackets${NC}"
+  echo -e "  \$ ${YELLOW}npm exec --package=@stdy/cli@0.22.1 -- steady path/to/your.openapi.yml --host 127.0.0.1 -p 4010 --validator-query-array-format=comma --validator-form-array-format=comma --validator-query-object-format=brackets --validator-form-object-format=brackets${NC}"
   echo
 
   exit 1

src/internal/utils/env.ts

@@ -9,10 +9,10 @@
  */
 export const readEnv = (env: string): string | undefined => {
   if (typeof (globalThis as any).process !== 'undefined') {
-    return (globalThis as any).process.env?.[env]?.trim() ?? undefined;
+    return (globalThis as any).process.env?.[env]?.trim() || undefined;
   }
   if (typeof (globalThis as any).Deno !== 'undefined') {
-    return (globalThis as any).Deno.env?.get?.(env)?.trim();
+    return (globalThis as any).Deno.env?.get?.(env)?.trim() || undefined;
   }
   return undefined;
 };

src/resources/beta/clusters/clusters.ts

@@ -76,9 +76,7 @@ export interface Cluster {
 
   control_plane_nodes: Array<Cluster.ControlPlaneNode>;
 
-  driver_version: 'CUDA_12_5_555' | 'CUDA_12_6_560' | 'CUDA_12_6_565' | 'CUDA_12_8_570';
-
-  duration_hours: number;
+  cuda_version: string;
 
   gpu_type: 'H100_SXM' | 'H200_SXM' | 'RTX_6000_PCI' | 'L40_PCIE' | 'B200_SXM' | 'H100_SXM_INF';
 
@@ -88,6 +86,8 @@ export interface Cluster {
 
   num_gpus: number;
 
+  nvidia_driver_version: string;
+
   region: string;
 
   /**
@@ -107,6 +107,20 @@ export interface Cluster {
     | 'Deleting';
 
   volumes: Array<Cluster.Volume>;
+
+  capacity_pool_id?: string;
+
+  created_at?: string;
+
+  duration_hours?: number;
+
+  install_traefik?: boolean;
+
+  reservation_end_time?: string;
+
+  reservation_start_time?: string;
+
+  slurm_shm_size_gib?: number;
 }
 
 export namespace Cluster {
@@ -142,6 +156,8 @@ export namespace Cluster {
     num_gpus: number;
 
     status: string;
+
+    instance_id?: string;
   }
 
   export interface Volume {
@@ -170,9 +186,10 @@ export interface ClusterListRegionsResponse {
 export namespace ClusterListRegionsResponse {
   export interface Region {
     /**
-     * List of supported identifiable driver versions available in the region.
+     * List of supported identifiable cuda/nvidia driver versions pairs available in
+     * the region.
      */
-    driver_versions: Array<string>;
+    driver_versions: Array<Region.DriverVersion>;
 
     /**
      * Identifiable name of the region.
@@ -182,27 +199,47 @@ export namespace ClusterListRegionsResponse {
     /**
      * List of supported identifiable gpus available in the region.
      */
-    supported_instance_types?: Array<string>;
+    supported_instance_types: Array<string>;
+  }
+
+  export namespace Region {
+    /**
+     * CUDA/NVIDIA driver versions pair available in the region to use in the create
+     * cluster request.
+     */
+    export interface DriverVersion {
+      /**
+       * CUDA driver version.
+       */
+      cuda_version: string;
+
+      /**
+       * NVIDIA driver version.
+       */
+      nvidia_driver_version: string;
+    }
   }
 }
 
 export interface ClusterCreateParams {
   /**
    * RESERVED billing types allow you to specify the duration of the cluster
    * reservation via the duration_days field. ON_DEMAND billing types will give you
-   * ownership of the cluster until you delete it.
+   * ownership of the cluster until you delete it. SCHEDULED_CAPACITY billing types
+   * allow you to reserve capacity for a scheduled time window. You must specify the
+   * reservation_start_time and reservation_end_time with this request.
    */
-  billing_type: 'RESERVED' | 'ON_DEMAND';
+  billing_type: 'RESERVED' | 'ON_DEMAND' | 'SCHEDULED_CAPACITY';
 
   /**
    * Name of the GPU cluster.
    */
   cluster_name: string;
 
   /**
-   * NVIDIA driver version to use in the cluster.
+   * CUDA version for this cluster. For example, 12.5
    */
-  driver_version: 'CUDA_12_5_555' | 'CUDA_12_6_560' | 'CUDA_12_6_565' | 'CUDA_12_8_570';
+  cuda_version: string;
 
   /**
    * Type of GPU to use in the cluster
@@ -215,12 +252,36 @@ export interface ClusterCreateParams {
    */
   num_gpus: number;
 
+  /**
+   * Nvidia driver version for this cluster. For example, 550. Only some combination
+   * of cuda_version and nvidia_driver_version are supported.
+   */
+  nvidia_driver_version: string;
+
   /**
    * Region to create the GPU cluster in. Usable regions can be found from
    * `client.clusters.list_regions()`
    */
   region: string;
 
+  /**
+   * Maximum number of GPUs to which the cluster can be auto-scaled up. This field is
+   * required if auto_scaled is true.
+   */
+  auto_scale_max_gpus?: number;
+
+  /**
+   * Whether GPU cluster should be auto-scaled based on the workload. By default, it
+   * is not auto-scaled.
+   */
+  auto_scaled?: boolean;
+
+  /**
+   * ID of the capacity pool to use for the cluster. This field is optional and only
+   * applicable if the cluster is created from a capacity pool.
+   */
+  capacity_pool_id?: string;
+
   /**
    * Type of cluster to create.
    */
@@ -231,11 +292,47 @@ export interface ClusterCreateParams {
    */
   duration_days?: number;
 
+  /**
+   * Whether automated GPU node failover should be enabled for this cluster. By
+   * default, it is disabled.
+   */
+  gpu_node_failover_enabled?: boolean;
+
+  /**
+   * Whether to install Traefik ingress controller in the cluster. This field is only
+   * applicable for Kubernetes clusters and is false by default.
+   */
+  install_traefik?: boolean;
+
+  /**
+   * Reservation end time of the cluster. This field is required for SCHEDULED
+   * billing to specify the reservation end time for the cluster.
+   */
+  reservation_end_time?: string;
+
+  /**
+   * Reservation start time of the cluster. This field is required for SCHEDULED
+   * billing to specify the reservation start time for the cluster. If not provided,
+   * the cluster will be provisioned immediately.
+   */
+  reservation_start_time?: string;
+
   /**
    * Inline configuration to create a shared volume with the cluster creation.
    */
   shared_volume?: ClusterCreateParams.SharedVolume;
 
+  /**
+   * Custom Slurm image for Slurm clusters.
+   */
+  slurm_image?: string;
+
+  /**
+   * Shared memory size in GiB for Slurm cluster. This field is required if
+   * cluster_type is SLURM.
+   */
+  slurm_shm_size_gib?: number;
+
   /**
    * ID of an existing volume to use with the cluster creation.
    */
@@ -275,6 +372,12 @@ export interface ClusterUpdateParams {
    * example, 8, 16 or 24
    */
   num_gpus?: number;
+
+  /**
+   * Timestamp at which the cluster should be decommissioned. Only accepted for
+   * prepaid clusters.
+   */
+  reservation_end_time?: string;
 }
 
 Clusters.Storage = Storage;

src/resources/beta/jig/jig.ts

@@ -147,7 +147,7 @@ export interface Deployment {
   /**
    * GPUType specifies the type of GPU requested (if any) for this deployment
    */
-  gpu_type?: 'h100-80gb' | ' a100-80gb';
+  gpu_type?: 'h100-80gb' | 'h100-40gb-mig' | 'b200-192gb';
 
   /**
    * HealthCheckPath is the HTTP path used for health checks of the application
@@ -445,7 +445,7 @@ export interface JigUpdateParams {
   /**
    * GPUType specifies the GPU hardware to use (e.g., "h100-80gb")
    */
-  gpu_type?: 'h100-80gb';
+  gpu_type?: 'h100-80gb' | 'h100-40gb-mig' | 'b200-192gb';
 
   /**
    * HealthCheckPath is the HTTP path for health checks (e.g., "/health"). Set to
@@ -615,7 +615,7 @@ export interface JigDeployParams {
   /**
    * GPUType specifies the GPU hardware to use (e.g., "h100-80gb").
    */
-  gpu_type: 'h100-80gb';
+  gpu_type: 'h100-80gb' | 'h100-40gb-mig' | 'b200-192gb';
 
   /**
    * Image is the container image to deploy from registry.together.ai.