Changelog

Added support for application/json requests on /v1/images/edits for GPT image models. JSON requests use images (and optional mask) with image_url or file_id references instead of multipart uploads.

We have optimized our inference stack for API customers and GPT-5.2 and GPT-5.2-Codex now run ~40% faster. Model and model weights are unchanged.

Announced Open Responses: an open-source spec for building multi-provider, interoperable LLM interfaces built on top of the original OpenAI Responses API.

Released gpt-5.2-codex to the Responses API. GPT-5.2-Codex is a version of GPT-5.2 optimized for agentic coding tasks in Codex or similar environments. Read more here.

Updated the gpt-realtime-mini and gpt-audio-mini slugs to point to the 2025-12-15 snapshots. If you need the previous model snapshots, use gpt-realtime-mini-2025-10-06 and gpt-audio-mini-2025-10-06.

Updated the sora-2 slug to point to sora-2-2025-12-08. If you need the previous model snapshot, use sora-2-2025-10-06.

Updated the gpt-4o-mini-tts and gpt-4o-mini-transcribe slugs to point to the 2025-12-15 snapshots. If you need the previous model snapshots, use gpt-4o-mini-tts-2025-03-20 and gpt-4o-mini-transcribe-2025-03-20. We currently recomend using gpt-4o-mini-transcribe over gpt-4o-transcribe for the best results.

Fixed an issue where gpt-image-1.5 and chatgpt-image-latest were incorrectly using high fidelity for image edits through /v1/images/edits, even when fidelity was explicitly set to low (the default).

Added gpt-image-1.5 and chatgpt-image-latest to the Responses API image generation tool.

Released gpt-image-1.5 and chatgpt-image-latest, our latest and most advanced models for image generation. Read more here.

Released four new dated audio snapshots. These updates deliver reliability, quality, and voice fidelity improvements for real-time, voice-driven applications. Read more here.

• gpt-realtime-mini-2025-12-15
• gpt-audio-mini-2025-12-15
• gpt-4o-mini-transcribe-2025-12-15
• gpt-4o-mini-tts-2025-12-15

This launch also includes support for Custom voices for eligible customers.

Released GPT-5.2, the newest flagship model in the GPT-5 model family. GPT-5.2 shows improvements over the previous GPT-5.1 in:

• General intelligence
• Instruction following
• Accuracy and token efficiency
• Multimodality—especially vision
• Code generation—especially front-end UI creation
• Tool calling and context management in the API
• Spreadsheet understanding and creation.

What's new in 5.2 is a new xhigh reasoning effort level, concise reasoning summaries, and new context management using compaction.

Released client-side compaction. For long-running conversations with the Responses API, you can use the /responses/compact endpoint to shrink the context you send with each turn.

Released gpt-5.1-codex-max to the Responses API. GPT-5.1-Codex is our most intelligent coding model optimized for long-horizon, agentic coding tasks. Read more here.

Added support for DTMF key presses in the Realtime API. You can now receive DTMF events while using a Realtime sideband connection. See docs here for more information.

Released GPT-5.1, the newest flagship model in the GPT-5 model family. GPT-5.1 is trained to be especially proficient in:

• Steerability and faster responses when less thinking's required
• Code generation and coding use cases
• Agentic workflows

Note that GPT-5.1 defaults to a new none reasoning setting for faster responses when less thinking's required—different from the previous medium default setting in GPT-5.

Released enhanced role-based access controls (RBAC). Role-based access control (RBAC) lets you decide who can do what across your organization and projects—both through the API and in the Dashboard.

Released gpt-5.1-codex and gpt-5.1-codex-mini to the Responses API. GPT-5.1-Codex is a version of GPT-5.1 optimized for agentic coding tasks in Codex or similar environments. Read more here.

Released extended prompt cache retention. Extended prompt cache retention keeps cached prefixes active for longer, up to a maximum of 24 hours. Extended Prompt Caching works by offloading the key/value tensors to GPU-local storage when memory is full, significantly increasing the storage capacity available for caching.

gpt-oss-safeguard-120b and gpt-oss-safeguard-20b are safety reasoning models built-upon gpt-oss. Read more here.

Released Enterprise Key Management (EKM). Enterprise Key Management (EKM) allows you to encrypt your customer content at OpenAI using keys managed by your own external Key Management System (KMS).

Released UK data residency.

Released several new features at OpenAI DevDay:

Released gpt-5-pro, a version of GPT-5 that uses more compute to think harder and provide consistently better answers.

Released gpt-realtime-mini and gpt-audio-mini for more cost-efficient speech to speech performance.

Released gpt-image-1-mini for more cost-efficient image generation and editing.

Launched v1/videos for rich, detailed, and dynamic video generation and remixing with our latest Sora 2 and Sora 2 Pro models.

Launched Agent Builder for visually creating custom multi-agent workflows.

Launched ChatKit, an embeddable chat interface for deploying agents.

Released Trace Evals, Datasets, and Prompt Optimization tools.

Evals: Released Third-Party Model Support.

Launched Service health dashboard.

Released IP allowlist. IP allowlisting restricts API access to only the IP addresses or ranges you specify.

Added support for image and file as a tool call output in Responses API.

Launched special-purpose model gpt-5-codex, built and optimized for use with the Codex CLI.

The OpenAI Realtime API is now generally available. Learn more in our Realtime API guide.

Added support for connectors to the Responses API. Connectors are OpenAI-maintained MCP wrappers for popular services like Google apps, Dropbox, and more that can be used to give model read access to data stored in those services.

Released the Conversations API, which allows you to create and manage long-running conversations with the Responses API. See the migration guide to see a side-by-side comparison and learn how to migrate from an Assistants API integration to Responses and Conversations.

Released GPT-5 family of models in the API, including gpt-5, gpt-5-mini, and gpt-5-nano.

Introduced the minimal reasoning effort value to optimize for fast responses in GPT-5 models (which support reasoning).

Introduced custom tool call type, which allows for freeform inputs to and outputs from the model when tool calling.

Launched support for Priority processing. Priority processing delivers significantly lower and more consistent latency compared to Standard processing while keeping pay-as-you-go flexibility.

Released o3-deep-research and o4-mini-deep-research, deep research variants of our o-series reasoning models optimized for deep analysis and research tasks. Learn more in the deep research guide.

Added support for async event handling with webhooks. Reduced and simplified pricing for the web search tool. Added support for the web search tool.

New reusable prompts are now available in the dashboard and Responses API. Via API, you can now reference templates created in the dashboard via the prompt parameter (with a prompt id, optional version) and supply dynamic variables that can include strings, images, or file inputs. Reusable prompts are not available in Chat Completions. Learn more.

Released o3-pro, a version of the o3 reasoning model that uses more compute to answer hard problems with better reasoning and consistency. Prices for the o3 model have also been reduced for all API requests, including batch and flex processing.

Added fine-tuning support with direct preference optimization for the models gpt-4.1-2025-04-14, gpt-4.1-mini-2025-04-14, and gpt-4.1-nano-2025-04-14.

New model snapshots available for gpt-4o-audio-preview and gpt-4o-realtime-preview. Released Agents SDK for TypeScript.

Added support for new built-in tools in the Responses API, including remote MCP servers and code interpreter. Learn more about tools.

Added support for using strict mode for tool schemas when using parallel tool calling with non-fine-tuned models. Added new schema features, including string validation for email and other patterns and specifying ranges for numbers and arrays.

Launched codex-mini-latest in the API, optimized for use with the Codex CLI.

Launched support for reinforcement fine-tuning. Learn about available fine-tuning methods. gpt-4.1-nano is now available for fine-tuning.

Launched support for Enhanced API Budget Alerts & Auto-recharge Limits.

Added a new image generation model, gpt-image-1. This model sets a new standard for image generation, with improved quality and instruction following.

Updated the Image Generation and Edit endpoints to support new parameters specific to the gpt-image-1 model.

Added two new o-series reasoning models, o3 and o4-mini. They set a new standard for math, science, and coding, visual reasoning tasks, and technical writing.

Launched Codex, our code generation CLI tool.

Added gpt-4.1, gpt-4.1-mini, and gpt-4.1-nano models to the API. These new models feature improved instruction following, coding, and a larger context window (up to 1M tokens). gpt-4.1 and gpt-4.1-mini are available for supervised fine-tuning. Announced deprecation of gpt-4.5-preview.

Added gpt-4o-mini-tts, gpt-4o-transcribe, gpt-4o-mini-transcribe, and whisper-1 models to the Audio API.

Released o1-pro, a version of the o1 reasoning model that uses more compute to answer hard problems with better reasoning and consistency.

Released several new models and tools and a new API for agentic workflows:

• Released the Responses API, a new API for creating and using agents and tools.
• Released a set of built-in tools for the Responses API: web search, file search, and computer use.
• Released the Agents SDK, an orchestration framework for designing, building, and deploying agents.
• Announced new models: gpt-4o-search-preview, gpt-4o-mini-search-preview, computer-use-preview.
• Announced plans to bring all Assistants API features to the easier to use Responses API, with an anticipated sunset date for Assistants in 2026 (after achieving full feature parity).

Added metadata field support to fine-tuning jobs.

Released a research preview of GPT-4.5—our largest and most capable chat model yet. GPT-4.5's high "EQ" and understanding of user intent make it better at creative tasks and agentic planning.

Launched the API Usage Dashboard Update. This update addresses requests for additional data filters, such as project selection, date picker, and fine-grained intervals. There’s also better support for viewing usage across different products and service tiers.

Introducing data residency in Europe. Read more here.

Launched o3-mini, a new small reasoning model that is optimized for science, math, and coding tasks.

Expanded access to o1 model. The o1 series of models are trained with reinforcement learning to perform complex reasoning.

Launched Admin API Key Rotations, enabling customers to programmatically rotate their admin api keys.

Updated Admin API Invites, enabling customers to programmatically invite users to projects at the same time they are invited to organizations.

Added new models for o1, gpt-4o-realtime, gpt-4o-audio and more.

Added WebRTC connection method for the Realtime API.

Added reasoning_effort parameter for o1 models.

Added developer message role for o1 model. Note that o1-preview and o1-mini do not support system or developer messages.

Launched Preference Fine-tuning using Direct Preference Optimization (DPO).

Launched beta SDKs for Go and Java. Learn more.

Added Realtime API support in the Python SDK.

Launched Usage API, enabling customers to programmatically query activities and spending across OpenAI APIs.

Released gpt-4o-2024-11-20, our newest model in the gpt-4o series.

Released Predicted Outputs, which greatly reduces latency for model responses where much of the response is known ahead of time. This is most common when regenerating the content of documents and code files with only minor changes.

Added five new voice types in the Realtime API and Chat Completions API.

Released new gpt-4o-audio-preview model for chat completions, which supports both audio inputs and outputs. Uses the same underlying model as the Realtime API.

Released several new features at OpenAI DevDay in San Francisco:

Realtime API: Build fast speech-to-speech experiences into your applications using a WebSockets interface.

Model distillation: Platform for fine-tuning cost-efficient models with your outputs from a large frontier model.

Image fine-tuning: Fine-tune GPT-4o with images and text to improve vision capabilities.

Evals: Create and run custom evaluations to measure model performance on specific tasks.

Prompt caching: Discounts and faster processing times on recently seen input tokens.

Generate in playground: Easily generate prompts, function definitions, and structured output schemas in the playground using the Generate button.

Released new omni-moderation-latest moderation model, which supports both images and text (for some categories), supports two new text-only harm categories, and has more accurate scores.

Released o1-preview and o1-mini, new large language models trained with reinforcement learning to perform complex reasoning tasks.

Assistants API now supports including file search results used by the file search tool, and customizing ranking behavior.

GA release for gpt-4o-2024-08-06 fine-tuning—all API users can now fine-tune the latest GPT-4o model.

Released dynamic model for chatgpt-4o-latest—this model will point to the latest GPT-4o model used by ChatGPT.

Launched Structured Outputs—model outputs now reliably adhere to developer supplied JSON Schemas.

Released gpt-4o-2024-08-06, our newest model in the gpt-4o series.

Launched Admin and Audit Log APIs, allowing customers to programmatically administer their organization and monitor changes using the audit logs. Audit logging must be enabled within settings.

Launched self-serve SSO configuration, allowing Enterprise customers on custom and unlimited billing to set up authentication against their desired IDP.

Launched fine-tuning for GPT-4o mini, enabling even higher performance for specific use cases.

Released GPT-4o mini, our affordable an intelligent small model for fast, lightweight tasks.

Released Uploads to upload large files in multiple parts.

Parallel function calling can be disabled in Chat Completions and the Assistants API by passing parallel_tool_calls=false.

.NET SDK launched in Beta.

Added support for file search customizations.

Added support for archiving projects . Only organization owners can access this functionality.

Added support for setting cost limits on a per-project basis for pay as you go customers.

Released GPT-4o in the API. GPT-4o is our fastest and most affordable flagship model.

Added support for image inputs to the Assistants API.

Added support for fine-tuned models to the Batch API .

Added stream_options: {"include_usage": true} parameter to the Chat Completions and Completions APIs. Setting this gives developers access to usage stats when using streaming.

Added a new endpoint to delete a message from a thread in the Assistants API.

Added a new function calling option tool_choice: "required" to the Chat Completions and Assistants APIs.

Added a guide for the Batch API and Batch API support for embeddings models

Introduced a series of updates to the Assistants API , including a new file search tool allowing up to 10,000 files per assistant, new token controls, and support for tool choice.

Introduced project based hierarchy for organizing work by projects, including the ability to create API keys and manage rate and cost limits on a per-project basis (cost limits available only for Enterprise customers).

Released Batch API

Released GPT-4 Turbo with Vision in general availability in the API

Added support for seed in the fine-tuning API

Added support for checkpoints in the fine-tuning API

Added support for adding Messages when creating a Run in the Assistants API

Added support for filtering Messages by run_id in the Assistants API

Added support for temperature and assistant message creation in the Assistants API

Added support for streaming in the Assistants API

Added timestamp_granularities parameter to the Audio API

Released gpt-3.5-turbo-0125, an updated GPT-3.5 Turbo model

Released embedding V3 models and an updated GPT-4 Turbo preview

Added dimensions parameter to the Embeddings API

Added additional_instructions parameter to run creation in the Assistants API

Added logprobs and top_logprobs parameters to the Chat Completions API

Changed function parameters argument on a tool call to be optional

Released OpenAI Deno SDK

Released GPT-4 Turbo Preview, updated GPT-3.5 Turbo, GPT-4 Turbo with Vision, Assistants API, DALL·E 3 in the API, and text-to-speech API

Deprecated the Chat Completions functions parameter in favor of tools

Released OpenAI Python SDK V1.0

Added encoding_format parameter to the Embeddings API

Added max_tokens to the Moderation models

Added function calling support to the Fine-tuning API