OpenAI Models
Explore the OpenAI language and embedding models available through our OpenAI Assistants API-compatible service.
OpenAI: GPT-4o-mini
- Context Length:
- 128,000 tokens
- Architecture:
- text+image->text
- Max Output:
- 16,384 tokens
Pricing:
GPT-4o Mini is OpenAI’s latest innovation, following the release of GPT-4 Omni. This cutting-edge model supports both text and image inputs, delivering text outputs with exceptional precision. Designed as a highly efficient small model, GPT-4o Mini offers state-of-the-art (SOTA) intelligence at a significantly reduced cost—making it over 60% more affordable than GPT-3.5 Turbo and a compelling choice compared to other leading models.
With an impressive 82% score on the MMLU benchmark, GPT-4o Mini currently outperforms GPT-4 on chat preference rankings, as reflected on common leaderboards. Its blend of advanced capabilities and cost-effectiveness positions it as a standout solution for multimodal applications.
For further details, explore the official launch announcement.
OpenAI: GPT-4o-mini (2024-07-18)
- Context Length:
- 128,000 tokens
- Architecture:
- text+image->text
- Max Output:
- 16,384 tokens
Pricing:
Introducing GPT-4o mini, OpenAI’s latest innovation following the groundbreaking GPT-4 Omni. This cutting-edge model supports both text and image inputs, delivering high-quality text outputs. Designed as their most advanced small-scale model, GPT-4o mini offers exceptional value, costing significantly less than other leading-edge models—over 60% more affordable than GPT-3.5 Turbo. Despite its compact size, it maintains state-of-the-art (SOTA) intelligence, making it a cost-efficient powerhouse.
GPT-4o mini achieves an impressive 82% score on MMLU (Massive Multitask Language Understanding) and currently outperforms GPT-4 on common leaderboards for chat preferences, solidifying its position as a top contender in AI performance.
To explore more about this revolutionary model, check out the official launch announcement.
OpenAI: GPT-3.5 Turbo 16k
- Context Length:
- 16,385 tokens
- Architecture:
- text->text
- Max Output:
- 4,096 tokens
Pricing:
Introducing the enhanced GPT-3.5 Turbo model, designed to deliver superior performance with advanced features such as improved instruction adherence, JSON mode for structured outputs, reproducible results, and parallel function calling. Trained on data up to September 2021, this iteration boasts higher accuracy in generating responses in specified formats and resolves a text encoding issue that previously affected non-English language function calls. Experience a more reliable and versatile AI solution tailored to meet your needs.
OpenAI: GPT-3.5 Turbo
- Context Length:
- 16,385 tokens
- Architecture:
- text->text
- Max Output:
- 4,096 tokens
Pricing:
Introducing GPT-3.5 Turbo, OpenAI’s most efficient and high-speed model. Designed to excel in both natural language understanding and generation, as well as code-related tasks, GPT-3.5 Turbo is expertly optimized for chat applications and traditional text completion. Its training data is current up to September 2021, ensuring reliable and up-to-date performance for a wide range of applications.
OpenAI: GPT-3.5 Turbo 16k (older v1106)
- Context Length:
- 16,385 tokens
- Architecture:
- text->text
- Max Output:
- 4,096 tokens
Pricing:
Introducing an enhanced version of the GPT-3.5 Turbo model, designed to deliver superior performance with advanced features. This iteration excels in instruction following, supports JSON mode for structured data handling, ensures reproducible outputs for consistent results, and enables parallel function calling for increased efficiency. Trained on data up to September 2021, this model combines reliability with cutting-edge capabilities to meet diverse needs.
OpenAI: GPT-3.5 Turbo (older v0613)
- Context Length:
- 4,095 tokens
- Architecture:
- text->text
- Max Output:
- 4,096 tokens
Pricing:
GPT-3.5 Turbo is OpenAI’s most efficient and high-speed model, designed to excel in both natural language understanding and generation, as well as code-related tasks. It is specifically optimized for chat applications and traditional text completion, delivering exceptional performance. The model’s training data is current up to September 2021, ensuring it leverages the latest advancements available at that time.
OpenAI: GPT-4o
- Context Length:
- 128,000 tokens
- Architecture:
- text+image->text
- Max Output:
- 16,384 tokens
Pricing:
Introducing GPT-4o ("o" for "omni"), OpenAI's cutting-edge AI model designed to handle both text and image inputs while delivering text outputs. Building on the intelligence of GPT-4 Turbo, GPT-4o operates at twice the speed and is 50% more cost-efficient. This advanced model also boasts significant enhancements in processing non-English languages and offers superior visual understanding capabilities.
During its benchmarking phase, GPT-4o was temporarily referred to as "im-also-a-good-gpt2-chatbot", showcasing its versatility and performance against other models.
OpenAI: GPT-4o (2024-08-06)
- Context Length:
- 128,000 tokens
- Architecture:
- text+image->text
- Max Output:
- 16,384 tokens
Pricing:
The 2024-08-06 version of GPT-4o introduces enhanced capabilities for structured outputs, including the ability to generate responses in a specified JSON schema using the response_format
parameter. Learn more about this feature here.
GPT-4o (where "o" stands for "omni") is OpenAI's most advanced AI model to date, capable of processing both text and image inputs while delivering text outputs. It matches the intelligence of GPT-4 Turbo but operates at twice the speed and is 50% more cost-effective. Additionally, GPT-4o demonstrates significant improvements in handling non-English languages and offers superior visual understanding.
During benchmarking, this model was temporarily referred to as "im-also-a-good-gpt2-chatbot", showcasing its competitive edge against other AI models.
OpenAI: GPT-4o (2024-11-20)
- Context Length:
- 128,000 tokens
- Architecture:
- text+image->text
- Max Output:
- 16,384 tokens
Pricing:
Introducing GPT-4o ("o" for "omni"), OpenAI’s cutting-edge AI model, released on November 20, 2024. This advanced version elevates creative writing with more natural, engaging, and tailored outputs, ensuring greater relevance and readability. GPT-4o also excels at handling uploaded files, delivering deeper insights and more comprehensive responses.
Building on the intelligence of GPT-4 Turbo, GPT-4o operates at twice the speed while being 50% more cost-effective. It supports both text and image inputs, generating text outputs with enhanced visual understanding. Additionally, GPT-4o boasts improved performance in processing non-English languages, making it a versatile and powerful tool for global users.
OpenAI: GPT-3.5 Turbo 16k
- Context Length:
- 16,385 tokens
- Architecture:
- text->text
- Max Output:
- 4,096 tokens
Pricing:
This advanced model provides four times the context length of gpt-3.5-turbo, enabling it to process approximately 20 pages of text in a single request. While this enhanced capability comes at a higher cost, it ensures greater efficiency for handling extensive inputs. The model's training data is current up to September 2021, ensuring it leverages the latest information available up to that point.
OpenAI: GPT-4o (2024-05-13)
- Context Length:
- 128,000 tokens
- Architecture:
- text+image->text
- Max Output:
- 4,096 tokens
Pricing:
Introducing GPT-4o ("o" for "omni"), OpenAI's cutting-edge AI model designed to deliver exceptional versatility and efficiency. GPT-4o seamlessly handles both text and image inputs, generating high-quality text outputs. It retains the advanced intelligence of GPT-4 Turbo while achieving remarkable improvements: doubling the processing speed and reducing costs by 50%. Additionally, GPT-4o excels in processing non-English languages and offers enhanced visual capabilities, making it a powerful tool for diverse applications.
During its benchmarking phase against other models, GPT-4o was temporarily referred to as "im-also-a-good-gpt2-chatbot", showcasing its robust performance and adaptability.
OpenAI: GPT-4o (extended)
- Context Length:
- 128,000 tokens
- Architecture:
- text+image->text
- Max Output:
- 64,000 tokens
Pricing:
Introducing GPT-4o ("o" for "omni"), OpenAI's cutting-edge AI model designed to handle both text and image inputs while delivering text outputs. Building on the intelligence of GPT-4 Turbo, GPT-4o operates at twice the speed and is 50% more cost-efficient. This advanced model also boasts superior performance in processing non-English languages and offers enhanced visual understanding capabilities.
During its benchmarking phase, GPT-4o was temporarily referred to as "im-also-a-good-gpt2-chatbot", showcasing its versatility and competitive edge.
OpenAI: GPT-4 Turbo
- Context Length:
- 128,000 tokens
- Architecture:
- text+image->text
- Max Output:
- 4,096 tokens
Pricing:
Introducing the advanced GPT-4 Turbo model, now enhanced with cutting-edge vision capabilities. This powerful update allows vision-based requests to leverage JSON mode and function calling, delivering greater flexibility and precision. Trained on data up to December 2023, GPT-4 Turbo combines the latest advancements in AI with unparalleled performance, making it an indispensable tool for innovative applications.
OpenAI: GPT-4 Turbo (older v1106)
- Context Length:
- 128,000 tokens
- Architecture:
- text->text
- Max Output:
- 4,096 tokens
Pricing:
Introducing the advanced GPT-4 Turbo, now enhanced with cutting-edge vision capabilities. This latest iteration supports vision-based requests through JSON mode and function calling, offering greater flexibility and precision. Trained on data up to April 2023, it delivers up-to-date, reliable, and intelligent responses tailored to your needs. Experience the next level of AI-powered innovation with GPT-4 Turbo.
OpenAI: GPT-4 Turbo Preview
- Context Length:
- 128,000 tokens
- Architecture:
- text->text
- Max Output:
- 4,096 tokens
Pricing:
Introducing the preview of GPT-4, a cutting-edge model designed to deliver enhanced instruction-following capabilities, JSON mode support, reproducible outputs, and parallel function calling, among other advanced features. Trained on data up to December 2023, this iteration represents a significant leap forward in AI performance and versatility.
Please note: During the preview phase, access to this model is subject to strict rate limits imposed by OpenAI.
OpenAI: GPT-4
- Context Length:
- 8,191 tokens
- Architecture:
- text->text
- Max Output:
- 4,096 tokens
Pricing:
OpenAI’s cutting-edge flagship model, GPT-4, represents a significant leap forward in artificial intelligence. As a large-scale multimodal language model, it excels at tackling complex challenges with enhanced precision, outperforming its predecessors through its expansive general knowledge and sophisticated reasoning abilities. GPT-4’s training data encompasses information up to September 2021, ensuring it draws from a comprehensive and up-to-date knowledge base to deliver accurate and insightful solutions.
OpenAI: GPT-4 (older v0314)
- Context Length:
- 8,191 tokens
- Architecture:
- text->text
- Max Output:
- 4,096 tokens
Pricing:
GPT-4-0314 marks the inaugural release of the GPT-4 model, featuring an impressive context length of 8,192 tokens. This version was actively supported until June 14, 2023, and its training data encompasses information up to September 2021, ensuring a robust and comprehensive knowledge base for its applications.
OpenAI: GPT-4 32k (older v0314)
- Context Length:
- 32,767 tokens
- Architecture:
- text->text
- Max Output:
- 4,096 tokens
Pricing:
GPT-4-32k is an advanced iteration of GPT-4, offering identical core capabilities but with a significantly expanded context window—four times larger than its predecessor. This enhancement enables the model to process up to 40 pages of text in a single interaction, making it exceptionally well-suited for managing extensive documents, such as PDFs, without the need for external vector databases. Trained on data up to September 2021, GPT-4-32k combines cutting-edge performance with the ability to handle complex, long-form content seamlessly.
OpenAI: GPT-4 32k
- Context Length:
- 32,767 tokens
- Architecture:
- text->text
- Max Output:
- 4,096 tokens
Pricing:
GPT-4-32k is an advanced iteration of GPT-4, offering identical core capabilities but with a significantly expanded context window—four times larger than its predecessor. This enhancement enables the model to process up to 40 pages of text in a single pass, making it exceptionally well-suited for managing lengthy documents, such as interacting with PDFs, without the need for external vector databases. The model's training data is current up to September 2021, ensuring it leverages the latest information available at that time.
Ready to build with OpenAI?
Start using these powerful models in your applications with our flexible pricing plans.