v0.11.0

Special thanks to @isKONSTANTIN for his first contribution in #17 🎉

This release contains the following additions:

General

• Made apiKey an optional parameter when building OpenAI
• Add support for structured outputs
• Add support for Administration endpoints

In order to use the Administration endpoints, admin key has to be configured when building OpenAI:

OpenAI openAI = OpenAI.newBuilder()
    .adminKey(System.getenv("OPENAI_ADMIN_KEY"))     
    .build();

Runs/RunSteps

• Add file search details and ranker options

Full Changelog: v0.10.0...v0.11.0

The artifact is available on Maven Central and can be imported as follows:

Gradle

implementation("io.github.stefanbratanov:jvm-openai:0.11.0")

Maven

<dependency>
    <groupId>io.github.stefanbratanov</groupId>
    <artifactId>jvm-openai</artifactId>
    <version>0.11.0</version>
</dependency>

v0.10.0

This release contains the following additions:

General

• Add support for Uploads
• Add GPT-4o mini model to OpenAIModel enum

Chat

• Add service_tier parameter when creating a chat completion

Full Changelog: v0.9.3...v0.10.0

The artifact is available on Maven Central and can be imported as follows:

Gradle

implementation("io.github.stefanbratanov:jvm-openai:0.10.0")

Maven

<dependency>
    <groupId>io.github.stefanbratanov</groupId>
    <artifactId>jvm-openai</artifactId>
    <version>0.10.0</version>
</dependency>

v0.9.3

This release contains the following additions:

Chat

• Ability to disable parallel function calling via parallel_tool_calls variable in CreateChatCompletionRequest

Messages

• Remove quote from FileCitation inside ThreadMessage

Runs

• Ability to disable parallel function calling via parallel_tool_calls variable in CreateRunRequest and CreateThreadAndRunRequest
• Add parallel_tool_calls to ThreadRun

Vector Stores

• Added support for file search customizations via ChunkingStrategy

Full Changelog: v0.9.2...v0.9.3

The artifact is available on Maven Central and can be imported as follows:

Gradle

implementation("io.github.stefanbratanov:jvm-openai:0.9.3")

Maven

<dependency>
    <groupId>io.github.stefanbratanov</groupId>
    <artifactId>jvm-openai</artifactId>
    <version>0.9.3</version>
</dependency>

v0.9.2

This release contains the following fixes/additions:

General

• Fix the visibility of ToolResources static initialization methods (#9)
• Fix serializing of classes which use @JsonSubTypes annotation (#10)
• Add GPT_4o to OpenAIModel enum
• Add Role, Purpose and Voice enums

Messages

• Add ability to reference an image URL in the content of a message via the content parameter when creating a message

Full Changelog: v0.9.1...v0.9.2

The artifact is available on Maven Central and can be imported as follows:

Gradle

implementation("io.github.stefanbratanov:jvm-openai:0.9.2")

Maven

<dependency>
    <groupId>io.github.stefanbratanov</groupId>
    <artifactId>jvm-openai</artifactId>
    <version>0.9.2</version>
</dependency>

v0.9.1

This release contains the following fixes/additions:

General

• Add workaround for record deserialization issues reported in #3 and #7

Chat

• Add ability to pass stream_options when creating a chat completion. More information for this functionality can be found at https://community.openai.com/t/usage-stats-now-available-when-using-streaming-with-the-chat-completions-api-or-completions-api/738156.
• Add usage field to ChatCompletionChunk

Full Changelog: v0.9.0...v0.9.1

The artifact is available on Maven Central and can be imported as follows:

Gradle

implementation("io.github.stefanbratanov:jvm-openai:0.9.1")

Maven

<dependency>
    <groupId>io.github.stefanbratanov</groupId>
    <artifactId>jvm-openai</artifactId>
    <version>0.9.1</version>
</dependency>

v0.9.0

This release has too many changes to mention all, but in short it provides support for Assistants API v2. If you would like to continue using v1, use Release v0.8.0 instead. A guide for migrating from v1 to v2 is available at https://platform.openai.com/docs/assistants/migration.

This release also contains the following additions/fixes:

General

• When adding parameters for a Function Tool, escaped JSON strings will be unescaped when serializing .

Runs

• Moved run steps functionality to a separate RunStepsClient to align with the spec: https://platform.openai.com/docs/api-reference/run-steps

Full Changelog: v0.8.0...v0.9.0

The artifact is available on Maven Central and can be imported as follows:

Gradle

implementation("io.github.stefanbratanov:jvm-openai:0.9.0")

Maven

<dependency>
    <groupId>io.github.stefanbratanov</groupId>
    <artifactId>jvm-openai</artifactId>
    <version>0.9.0</version>
</dependency>

v0.8.0

This release contains the following additions/fixes:

General

• Add support for Batch API
• Add ability to pass a project id (OpenAI-Project header) when building the OpenAI instance

Fine-tuning

• Change learning_rate_multiplier parameter in Hyperparameters to be a double instead of an int

Assistants

• Add ability to pass temperature, top_p and response_format when creating an assistant

Runs

• Add ability to pass top_p, max_prompt_tokens, max_completion_tokens, truncation_strategy, tool_choice and response_format when creating a run
• Add top_p, max_prompt_tokens, max_completion_tokens, truncation_strategy, tool_choice and response_format fields to ThreadRun

Full Changelog: v0.7.0...v0.8.0

The artifact is available on Maven Central and can be imported as follows:

Gradle

implementation("io.github.stefanbratanov:jvm-openai:0.8.0")

Maven

<dependency>
    <groupId>io.github.stefanbratanov</groupId>
    <artifactId>jvm-openai</artifactId>
    <version>0.8.0</version>
</dependency>

v0.7.0

Special thanks to @maciej-cz for his first contribution in #5 🎉

This release contains the following additions/improvements:

General

• Add OpenAIModel enum to simplify the selection of a model

Chat

• Improved handling of streaming events when ChatCompletionStreamSubscriber is used
• Renamed StreamChatCompletionSubscriber to ChatCompletionStreamSubscriber

Fine-tuning

• Add ability to pass integrations and seed fields when creating a fine-tuning job
• Add integrations and seed field to FineTuningJob
• Add support for listing checkpoints for a fine-tuning job.

Messages

• Add ability to pass run_id query param when listing messages to allow filtering by the run id that generated them

Runs

• Add ability to pass additional_messages, temperature and stream fields when creating a run
• Add temperature field to ThreadRun
• Add ability to pass stream field when submitting tool outputs to run
• Add createRunAndStream, createThreadAndRunAndStream and submitToolOutputsAndStream to RunsClient to support Assistants Streaming

Full Changelog: v0.6.2...v0.7.0

The artifact is available on Maven Central and can be imported as follows:

Gradle

implementation("io.github.stefanbratanov:jvm-openai:0.7.0")

Maven

<dependency>
    <groupId>io.github.stefanbratanov</groupId>
    <artifactId>jvm-openai</artifactId>
    <version>0.7.0</version>
</dependency>

v0.6.2

This release contains the following additions/improvements:

General

• Upgraded Jackson library to version 2.17.0
• Removed argument validations when building requests, which allows the library to be more easily maintained in case of API changes

Full Changelog: v0.6.1...v0.6.2

The artifact is available on Maven Central and can be imported as follows:

Gradle

implementation("io.github.stefanbratanov:jvm-openai:0.6.2")

Maven

<dependency>
    <groupId>io.github.stefanbratanov</groupId>
    <artifactId>jvm-openai</artifactId>
    <version>0.6.2</version>
</dependency>

v0.6.1

This release contains the following additions/improvements:

General

• Add ability to set a requestTimeout when initializing OpenAI. This timeout will apply to all requests.

OpenAI openAI = OpenAI.newBuilder(System.getenv("OPENAI_API_KEY"))
    .requestTimeout(Duration.ofSeconds(10))
    .build();

Chat

• Remove instance_id parameter as per openai/openai-openapi#198

Full Changelog: v0.6.0...v0.6.1

The artifact is available on Maven Central and can be imported as follows:

Gradle

implementation("io.github.stefanbratanov:jvm-openai:0.6.1")

Maven

<dependency>
    <groupId>io.github.stefanbratanov</groupId>
    <artifactId>jvm-openai</artifactId>
    <version>0.6.1</version>
</dependency>