> ## Documentation Index > Fetch the complete documentation index at: https://docs.together.ai/llms.txt > Use this file to discover all available pages before exploring further. # List supported models > Returns the models supported by the RL service and their limits for training/sampling operations. ## OpenAPI ````yaml openapi.yaml GET /rl/supported-models openapi: 3.1.0 info: title: Together APIs description: The Together REST API. See https://docs.together.ai for more details. version: 2.0.0 termsOfService: https://www.together.ai/terms-of-service contact: name: Together Support url: https://www.together.ai/contact license: name: MIT url: https://github.com/togethercomputer/openapi/blob/main/LICENSE servers: - url: https://api.together.ai/v1 description: Default environment for APIs - url: https://api-inference.together.ai/v2 description: Optimized environment for inference security: - bearerAuth: [] paths: /rl/supported-models: get: tags: - RL summary: List supported models description: >- Returns the models supported by the RL service and their limits for training/sampling operations. operationId: listSupportedModels responses: '200': description: List of supported RL models content: application/json: schema: $ref: '#/components/schemas/RL.SupportedModelsListResponse' '429': description: Rate limit exceeded. content: application/json: schema: $ref: '#/components/schemas/ErrorData' default: description: An unexpected error response. content: application/json: schema: $ref: '#/components/schemas/ErrorData' components: schemas: RL.SupportedModelsListResponse: type: object description: List of base models supported by the RL service required: - data properties: data: type: array items: $ref: '#/components/schemas/RL.SupportedModel' description: Supported base models for RL ErrorData: type: object required: - error properties: error: type: object properties: message: type: string nullable: false type: type: string nullable: false param: type: string nullable: true default: null code: type: string nullable: true default: null required: - type - message RL.SupportedModel: type: object description: >- A base model supported by the RL service. Per-mode configs are present only when the model supports that mode. required: - base_model - default_gpu_type properties: base_model: type: string example: Qwen/Qwen3.5-4B description: >- Base model identifier to pass as base_model when creating a model resource default_gpu_type: type: string example: H100-80GB enum: - H100-80GB - B200-SXM description: GPU type used when model-resource creation omits gpu_type. compute_configs: type: array items: $ref: '#/components/schemas/RL.SupportedModelComputeConfig' description: Validated GPU configurations available for this base model. RL.SupportedModelComputeConfig: type: object description: A validated hardware configuration available for an RL base model. required: - gpu_type properties: gpu_type: type: string example: H100-80GB enum: - H100-80GB - B200-SXM description: GPU type this configuration provisions. trainer_config: $ref: '#/components/schemas/RL.ModelTrainerConfig' description: >- Training config for this GPU type. Set when the model supports at least one training mode on this GPU type. generator_config: $ref: '#/components/schemas/RL.ModelGeneratorConfig' description: >- Inference config for this GPU type. Set when the model can be provisioned with generator replicas on this GPU type. RL.ModelTrainerConfig: type: object description: >- Training capabilities and limits for a model on one GPU type. Each sub-mode is present only when the model supports that training mode. properties: full: $ref: '#/components/schemas/RL.ModelFullModeConfig' description: >- Full-weight training config. Set when the model supports full-weight training. lora: $ref: '#/components/schemas/RL.ModelLoraModeConfig' description: LoRA training config. Set when the model supports LoRA training. RL.ModelGeneratorConfig: type: object description: >- Inference capabilities and limits for a model on one GPU type. If present, the model can be provisioned with one or more generator replicas. required: - context_length - sampling_defaults properties: context_length: type: integer example: 40960 description: Maximum tokens in a single inference request (prompt + completion) sampling_defaults: $ref: '#/components/schemas/RL.ModelSamplingDefaults' description: Default sampling parameters used for sample requests. RL.ModelFullModeConfig: type: object description: >- Full-weight training config. If present, the model supports full-weight training. required: - max_batch_size - max_seq_length properties: max_batch_size: type: integer example: 512 description: Maximum global batch size accepted by a forward-backward step max_seq_length: type: integer example: 40960 description: Maximum sequence length in tokens RL.ModelLoraModeConfig: type: object description: LoRA training config. If present, the model supports LoRA training. required: - max_batch_size - max_seq_length - max_rank properties: max_batch_size: type: integer example: 512 description: Maximum global batch size accepted by a forward-backward step max_seq_length: type: integer example: 40960 description: Maximum sequence length in tokens max_rank: type: integer example: 64 description: Maximum LoRA rank RL.ModelSamplingDefaults: type: object description: Default sampling parameters used for a Sample request. required: - 'n' - max_tokens - temperature - logprobs properties: 'n': type: integer example: 1 description: Number of completions per prompt max_tokens: type: integer example: 512 description: Maximum tokens generated per completion temperature: type: number example: 1 description: Sampling temperature logprobs: type: integer example: 1 description: Number of logprobs to return per token securitySchemes: bearerAuth: type: http scheme: bearer x-bearer-format: bearer x-default: default ```` This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.