Skip to content
Hayonix

API reference · Fine-tuning

Get a fine-tuned model of the project

GET/v1/fine-tuned-models/{fine_tuned_model_id}
API keyfine_tune_job:readProject
TypeScript
client.fineTuning.getModel()
Python
client.fine_tuning.get_model()
CLI
hayonix fine-tunes model

Its artefacts, the job and checkpoint it came from, and its serving status: poll serving.shared.status after enabling it until it is ready.

Project API key (Authorization: Bearer sk_…), or a console session. Requires the fine_tune_job:read permission (a key's scope, or a role that includes it).

Examples

The samples read API_KEY and API_BASE_URL from the environment (see Setup). Values are examples.

curl -sS "$API_BASE_URL/fine-tuned-models/ftm_01jc5zq4m8k3v6x9t2w7r0b1e4" \
  -H "Authorization: Bearer $API_KEY"

Request

Headers
HeaderRequiredDescription
AuthorizationyesBearer and your API key. The key implies its project.
X-Project-Idsession onlyThe project to act in, for console sessions. Requests with an API key leave it out.

Path parameters

  • fine_tuned_model_idstringrequired

    ID of the fine-tuned model

    ID with prefix ftm_

Response

200 FineTunedModel — The model

  • idstringrequired

    Unique ID (`ftm_…`)

    ID with prefix ftm_

  • objectstringrequired

    Always `fine_tuned_model`: the type of this object

    One offine_tuned_model

  • project_idstringrequired

    Project it belongs to

    ID with prefix prj_

  • namestringrequired

    The model name, unique in the org

  • fine_tune_idstringrequired

    ID of the fine-tuning job

    ID with prefix ftj_

  • checkpoint_idstringrequired

    ID of the checkpoint

    ID with prefix ftc_

  • base_modelobjectrequired

    The model that was fine-tuned

    Show 2 fields of base_model
    • idstringrequired

      Unique ID (`mdl_…`)

      ID with prefix mdl_

    • slugstringrequired

      The base model's name (`publisher/name`)

  • methodstringrequired

    The training method of the job that produced it

    One ofsupervisedpreference

  • adapter_typestringrequired

    `lora`: a low-rank adapter on the base model; `full`: full weights

    One oflorafull

  • adapterobject | nullrequired

    The LoRA adapter; null for a full fine-tune

    Show 6 fields of adapter
    • formatstringrequired

      File format of the adapter

    • bytesintegerrequired

      Size in bytes

    • filesarray of objectrequired

      The files that make up the adapter

      Show 3 fields of files
      • pathstringrequired

        Path of the file inside the adapter

      • bytesintegerrequired

        Size in bytes

      • sha256stringrequired

        SHA-256 of the content, hex

    • rankinteger | nullrequired

      LoRA rank; null when not recorded

    • alphainteger | nullrequired

      LoRA alpha; null when not recorded

    • target_modulesstring | array of string | any JSON | nullrequired

      Layers the adapter was trained on (`all-linear` or module names); null: the model's usual LoRA targets

  • merged_weightsobject | nullrequired

    Full weights: the base model with the adapter merged in, or the result of a full fine-tune; null when a LoRA job was created with `merge_weights: false`

    Show 3 fields of merged_weights
    • formatstringrequired

      File format of the weights

    • bytesintegerrequired

      Size in bytes

    • filesarray of objectrequired

      The files that make up the weights

      Show 3 fields of files
      • pathstringrequired

        Path of the file inside the weights

      • bytesintegerrequired

        Size in bytes

      • sha256stringrequired

        SHA-256 of the content, hex

  • catalog_model_idstring | nullrequired

    ID of the catalog model

    ID with prefix mdl_

  • servingobjectrequired

    How the model can be served: per token on the shared pool of its base model (`shared`), or on a dedicated endpoint (create one with `model.source = fine_tuned_model`)

    Show 4 fields of serving
    • availablebooleanrequired

      Callable per token right now under `name` (shared serving is ready)

    • sharedobjectrequired

      Shared (per-token) serving

      Show 6 fields of shared
      • statestringrequired

        Per-token serving on the shared pool of the base model, as set by enable / disable (LoRA only)

        One ofdisabledenabled

      • statusstringrequired

        disabled; loading (enabled, the pool is starting replicas with the adapter); ready (callable)

        One ofdisabledloadingready

      • reasonstring | nullrequired

        Why the state last changed: requested, or pool_removed (by the platform)

      • changed_atstring (date-time) | nullrequired

        Time changed (ISO 8601, UTC)

      • priceobject | nullrequired

        What calling the model per token costs (nano-USD per 1M tokens, the fine-tuned price of its base model in the price book in force); null when fine-tuned inference on the base is not priced

        Show 6 fields of price
        • currencystringrequired

          Always `usd`

          One ofUSD

        • unitstringrequired

          Every price here is nano-USD per 1M tokens

          One ofnano_usd_per_1m_tokens

        • inputstringrequired

          Integer nano-USD as a decimal string

          pattern ^-?[0-9]+$

        • cached_inputstringrequired

          Integer nano-USD as a decimal string

          pattern ^-?[0-9]+$

        • outputstringrequired

          Integer nano-USD as a decimal string

          pattern ^-?[0-9]+$

        • batchobject | nullrequired

          The batch API price for this model, when priced

      • blockedobject | nullrequired

        While disabled: why `enable` would be refused right now (the same code it would answer with), or null when it would be accepted (the `shared_adapters` quota is checked only by `enable`)

        Show 2 fields of blocked
        • codestringrequired

          Machine-readable code

          One ofnot_an_adapterno_shared_pooladapter_format_unavailableadapter_rank_unsupportednot_pricedshared_pool_full

        • messagestringrequired

          One sentence for people

    • dedicatedobjectrequired

      Whether a dedicated endpoint deployment on each engine finds weights for this model (false: creating one answers `artifact_unavailable`; llama.cpp needs the merged GGUF, which older models lack)

      Show 2 fields of dedicated
      • llama_cppbooleanrequired

        A llama.cpp deployment finds weights for this model

      • vllmbooleanrequired

        A vLLM deployment finds weights for this model

    • formatsarray of stringrequired

      Serving formats the model has besides its safetensors: adapter_gguf, merged_gguf (llama.cpp engines)

  • created_atstring (date-time)required

    When it was created (ISO 8601, UTC)

Example response
Example response
{
  "id": "ftm_01jc5zq4m8k3v6x9t2w7r0b1e4",
  "object": "fine_tuned_model",
  "project_id": "prj_01jc5zq4m8k3v6x9t2w7r0b1e4",
  "name": "string",
  "fine_tune_id": "ftj_01jc5zq4m8k3v6x9t2w7r0b1e4",
  "checkpoint_id": "ftc_01jc5zq4m8k3v6x9t2w7r0b1e4",
  "base_model": {
    "id": "mdl_01jc5zq4m8k3v6x9t2w7r0b1e4",
    "slug": "string"
  },
  "method": "supervised",
  "adapter_type": "lora",
  "adapter": {
    "format": "string",
    "bytes": 1,
    "files": [
      {
        "path": "string",
        "bytes": 1,
        "sha256": "string"
      }
    ],
    "rank": 1,
    "alpha": 1,
    "target_modules": "string"
  },
  "merged_weights": {
    "format": "string",
    "bytes": 1,
    "files": [
      {
        "path": "string",
        "bytes": 1,
        "sha256": "string"
      }
    ]
  },
  "catalog_model_id": "mdl_01jc5zq4m8k3v6x9t2w7r0b1e4",
  "serving": {
    "available": true,
    "shared": {
      "state": "disabled",
      "status": "disabled",
      "reason": "string",
      "changed_at": "2026-01-01T00:00:00.000Z",
      "price": {
        "currency": "USD",
        "unit": "nano_usd_per_1m_tokens",
        "input": "1500000000",
        "cached_input": "1500000000",
        "output": "1500000000",
        "batch": {
          "input": "1500000000",
          "cached_input": "1500000000",
          "output": "1500000000"
        }
      },
      "blocked": {
        "code": "not_an_adapter",
        "message": "string"
      }
    },
    "dedicated": {
      "llama_cpp": true,
      "vllm": true
    },
    "formats": [
      "string"
    ]
  },
  "created_at": "2026-01-01T00:00:00.000Z"
}

Errors

Errors use the standard envelope. What this operation can answer:

Errors
StatusTypeWhen
400invalid_request_errorA parameter or body field is missing or not allowed; param names it.
401authentication_errorThe key or session is missing, revoked or expired.
403permission_errorThe credential lacks the permission this needs, or the organization is suspended.
404not_found_errorNo such resource in this project (an ID from another project is answered the same way).
429rate_limit_error rate_limitedToo many requests; retry after retry-after seconds.