HeatWave User Guide  /  ...  /  Supported LLMs, Embedding Models, and Languages

4.2.2 Supported LLMs, Embedding Models, and Languages

This topic provides the list of languages that HeatWave GenAI supports and the embedding models as well as large language models (LLMs) that are available in HeatWave.

This topic contains the following sections:

Viewing Available Models

As of MySQL 9.3.2, once you are connected to the DB System, you can view the list of available models as shown below:

mysql> SELECT * FROM sys.ML_SUPPORTED_LLMS;

The output is similar to the following:

+---------------------------+--------------------------------------+-------------------+---------------------+---------------+
| provider                  | model_id                             | availability_date | capabilities        | default_model |
+---------------------------+--------------------------------------+-------------------+---------------------+---------------+
| HeatWave                  | llama2-7b-v1                         | 2024-07-01        | ["GENERATION"]      |             0 |
| HeatWave                  | llama3-8b-instruct-v1                | 2024-07-24        | ["GENERATION"]      |             1 |
| HeatWave                  | llama3.1-8b-instruct-v1              | 2025-05-20        | ["GENERATION"]      |             0 |
| HeatWave                  | llama3.2-1b-instruct-v1              | 2025-05-20        | ["GENERATION"]      |             0 |
| HeatWave                  | llama3.2-3b-instruct-v1              | 2025-05-20        | ["GENERATION"]      |             1 |
| HeatWave                  | mistral-7b-instruct-v1               | 2024-07-01        | ["GENERATION"]      |             0 |
| HeatWave                  | mistral-7b-instruct-v3               | 2025-05-20        | ["GENERATION"]      |             0 |
| HeatWave                  | all_minilm_l12_v2                    | 2024-07-01        | ["TEXT_EMBEDDINGS"] |             0 |
| HeatWave                  | multilingual-e5-small                | 2024-07-24        | ["TEXT_EMBEDDINGS"] |             1 |
| OCI Generative AI Service | meta.llama-3.3-70b-instruct          | 2025-01-15        | ["GENERATION"]      |             0 |
| OCI Generative AI Service | meta.llama-3.2-11b-vision-instruct   | 2024-11-13        | ["GENERATION"]      |             0 |
| OCI Generative AI Service | cohere.command-r-08-2024             | 2024-11-13        | ["GENERATION"]      |             0 |
| OCI Generative AI Service | meta.llama-3.2-90b-vision-instruct   | 2024-11-13        | ["GENERATION"]      |             0 |
| OCI Generative AI Service | cohere.command-r-plus-08-2024        | 2024-11-01        | ["GENERATION"]      |             0 |
| OCI Generative AI Service | meta.llama-3.1-70b-instruct          | 2024-09-18        | ["GENERATION"]      |             0 |
| OCI Generative AI Service | meta.llama-3.1-405b-instruct         | 2024-09-18        | ["GENERATION"]      |             0 |
| OCI Generative AI Service | cohere.embed-english-light-v3.0      | 2023-11-30        | ["TEXT_EMBEDDINGS"] |             0 |
| OCI Generative AI Service | cohere.embed-english-v3.0            | 2023-11-30        | ["TEXT_EMBEDDINGS"] |             0 |
| OCI Generative AI Service | cohere.embed-multilingual-light-v3.0 | 2023-11-30        | ["TEXT_EMBEDDINGS"] |             0 |
| OCI Generative AI Service | cohere.embed-multilingual-v3.0       | 2023-11-30        | ["TEXT_EMBEDDINGS"] |             0 |
| OCI Generative AI Service | cohere.embed-english-light-v2.0      | 2023-09-08        | ["TEXT_EMBEDDINGS"] |             0 |
+---------------------------+--------------------------------------+-------------------+---------------------+---------------+

However, to view or use the Oracle Cloud Infrastructure (OCI) Generative AI service models in this list, you need to enable the database system to access OCI services. For more information, see Section 4.2.3, “Authenticating OCI Generative AI Service”.

HeatWave In-Database LLMs

The following in-database LLMs are available in HeatWave:

  • As of MySQL 9.3.1:

    • llama3.2-3b-instruct-v1: supports all available Languages. This is the default LLM used by HeatWave GenAI for content generation.

    • llama3.2-1b-instruct-v1: supports all available Languages.

    • llama3.1-8b-instruct-v1: supports all available Languages.

    • mistral-7b-instruct-v3: supports all available Languages.

  • As of MySQL 9.0.1-u1:

    • llama3-8b-instruct-v1: supports English (en), French (fr), German (de), Hindi (hi), Italian (it), Portuguese (pt), Spanish (es), and Thai (th).

  • As of MySQL 9.0.0:

    • mistral-7b-instruct-v1: supports English (en).

      As of MySQL 9.0.1-u1, mistral-7b-instruct-v1 supports English (en), French (fr), German (de), and Spanish (es).

    • llama2-7b-v1: supports English (en).

      As of MySQL 9.0.1-u1, the support for llama2-7b-v1 has been deprecated.

As of MySQL 9.3.1, the in-database LLMs are included in all HeatWave Cluster shapes. However, the availability of LLMs depends on the HeatWave Cluster shape that you are using. The HeatWave.512GB shape supports all the in-database LLMs. Whereas, the HeatWave.Free and HeatWave.32GB shapes support llama3.2-3b-instruct-v1 and llama3.2-1b-instruct-v1 only.

In earlier versions of MySQL, the LLMs are included only in the HeatWave.512GB shape.

OCI Generative AI Service LLMs

HeatWave GenAI lets you use OCI Generative AI service LLMs, which support all available Languages.

However, to use the OCI Generative AI service with HeatWave GenAI, you need to enable the database system to access OCI services. For more information, see Section 4.2.3, “Authenticating OCI Generative AI Service”.

Note

These LLMs are available only in the regions supported by the OCI Generative AI service. For more information, see Pretrained Foundational Models in Generative AI.

As of MySQL 9.3.2, use the inbuilt model explorer to view the full list of available LLMs. This is because the list of OCI Generative AI service LLMs available in HeatWave might change before a new version of MySQL is released.

In previous versions of MySQL, the following OCI Generative AI service LLMs are available in HeatWave:

  • As of MySQL 9.2.2:

    • meta.llama-3.2-90b-vision-instruct

    • meta.llama-3.3-70b-instruct

  • As of MySQL 9.1.2:

    • cohere.command-r-08-2024

    • cohere.command-r-plus-08-2024

    • meta.llama-3.1-405b-instruct

To learn about deprecation and removal times of OCI Generative AI service LLMs, see Retiring the Models.

As of MySQL 9.3.1, using OCI Generative AI Service LLMs is supported for all HeatWave Cluster shapes. In earlier versions of MySQL, the using OCI Generative AI Service LLMs is supported only for the HeatWave.512GB shape.

HeatWave In-Database Embedding Models

The following in-database embedding models are available in HeatWave:

  • multilingual-e5-small:

    • As of MySQL 9.3.0, HeatWave GenAI uses this embedding model, by default, for encoding text or documents in all supported languages.

      In earlier versions of MySQL, HeatWave GenAI uses this embedding model, by default, for encoding text or documents in all supported languages other than English.

    • As of MySQL 9.2.1, can be used for encoding text or documents in all supported languages.

      In earlier versions of MySQL, can be used for encoding text or documents in all supported languages other than English.

    • This embedding model is available as of MySQL 9.0.1-u1.

  • minilm:

    • In MySQL 9.2.2 and earlier versions, HeatWave GenAI uses this embedding model, by default, for encoding English text or documents.

    • As of MySQL 9.2.1, can be used for encoding text or documents in any supported language.

      In earlier versions of MySQL, can be used for encoding text or documents in English only.

    all_minilm_l12_v2 is an alternative ID for the same model.

OCI Generative AI Service Embedding Models

OCI Generative AI service embedding models are available in HeatWave as of MySQL 9.0.1-u1.

Note

These models are available only in the regions supported by the OCI Generative AI service. For more information, see Pretrained Foundational Models in Generative AI.

However, to use the OCI Generative AI service with HeatWave GenAI, you need to enable the database system to access OCI services. For more information, see Section 4.2.3, “Authenticating OCI Generative AI Service”.

As of MySQL 9.3.2, use the inbuilt model explorer to view the full list of available embedding models. This is because the list of OCI Generative AI service embedding models available in HeatWave might change before a new version of MySQL is released.

In previous versions of MySQL, the following OCI Generative AI service embedding models are available in HeatWave:

  • cohere.embed-english-v3.0:

    • As of MySQL 9.2.1, can be used for encoding text or files in any supported language.

      In earlier versions of MySQL, can be used for encoding text or files in English only.

  • cohere.embed-multilingual-v3.0:

    • As of MySQL 9.2.1, can be used for encoding text or files in any supported language.

      In earlier versions of MySQL, can be used for encoding text or files in supported languages other than English.

Languages

As of MySQL 9.0.1-u1, HeatWave GenAI supports natural-language communication, ingesting documents, as well as generating text-based content in multiple languages. The quality of the generated text outputs depends on the training and ability of the LLM to work with the language.

As of MySQL 9.0.0, HeatWave GenAI supports natural-language communication, ingesting documents, as well as generating text-based content in English (en).

Following is a combined list of languages supported by HeatWave In-Database LLMs and OCI Generative AI Service LLMs:

  • Arabic (ar)

  • Bengali (bn)

  • Burmese (my)

  • Chinese (zh)

  • Czech (cs)

  • Dutch (nl)

  • English (en)

  • French (fr)

  • German (de)

  • Hebrew (he)

  • Hindi (hi)

  • Indonesian (id)

  • Italian (it)

  • Japanese (ja)

  • Khmer (km)

  • Korean (ko)

  • Lao (lo)

  • Malay (ms)

  • Persian (fa)

  • Polish (pl)

  • Portuguese (pt)

  • Spanish (es)

  • Tagalog (tl)

  • Thai (th)

  • Turkish (tr)

  • Urdu (ur)

  • Vietnamese (vi)

However, not all LLMs support all the languages. To learn which LLM supports which language, see HeatWave In-Database LLMs and OCI Generative AI Service LLMs.

Note

To set the value of the language parameter in HeatWave GenAI routines that support this parameter, do not use the language name to specify the language. Use the two-letter ISO 639-1 code for the language instead. For example, to use French, use the ISO 639-1 code for French, which is fr.