OpenAI's most capable and aligned model yet. Better at coding, instruction following, and with an extended 1 million token context window.
54.6% score on SWE-bench Verified, a 21.4% improvement over GPT-4o. Produces better code edits with minimal changes.
Scores 87.4% on the IFEval benchmark, compared to 81.0% for GPT-4o. Excels at handling complex multi-step instructions.
Process up to 1 million tokens in a single prompt, enabling analysis of extensive documents, code repositories, and long conversations.
Discover how GPT-4.1 sets new standards in AI with improved coding, instruction following, and context handling.
GPT-4.1 brings significant improvements to coding tasks, making it the most capable AI coding assistant available.
"Human graders preferred GPT-4.1 based websites over GPT-4o nearly 80% of the time in head-to-head comparisons." — OpenAI official evaluation
GPT-4.1 demonstrates remarkable progress in understanding and executing complex instructions with precision.
"On complex multi-step tasks, GPT-4.1 consistently outperforms previous models by accurately following instructions while maintaining context awareness." — Scale AI evaluation report
GPT-4.1 introduces groundbreaking advancements in handling extended contexts, unlocking new possibilities for document analysis.
"Thomson Reuters reported a 17% improvement in multi-document review accuracy, while Carlyle extracted granular financial data 50% faster than with previous models." — Industry case studies
The GPT-4.1 family includes three models tailored for different use cases and requirements.
The flagship model with the highest capabilities across all tasks.
Balanced performance and efficiency for everyday tasks.
The fastest and most economical option in the series.
See how GPT-4.1 stacks up against previous models in standardized benchmarks and real-world tests.
Higher scores indicate better performance on software engineering tasks.
Higher scores indicate better ability to follow complex instructions accurately.
60% improvement over GPT-4o in code modification benchmarks, resulting in quicker acceptance of code changes.
55% better code review suggestions when evaluated using GitHub pull requests compared to previous models.
53% increase in accuracy on challenging tax scenario benchmarks, improving research for tax professionals.
Nearly 2x improvement in complex SQL evaluation tasks, significantly reducing manual debugging time.
GPT-4.1 offers improved performance at lower costs, with flexible options for every need.
| Model | Input (per 1M tokens) | Cached Input | Output (per 1M tokens) | Blended Price* |
|---|---|---|---|---|
| GPT-4.1 | $2.00 | $0.50 | $8.00 | $1.84 |
| GPT-4.1 mini | $0.40 | $0.10 | $1.60 | $0.42 |
| GPT-4.1 nano | $0.10 | $0.025 | $0.40 | $0.12 |
*Blended price based on typical input/output and caching ratios.
GPT-4.1 is 26% less expensive than GPT-4o for median queries. The nano model is the cheapest and fastest in the series.
75% discount for cached prompts (up from 50%), drastically improving efficiency for repetitive tasks and high-volume applications.
Use the full 1 million token context with no additional charges beyond the standard per-token pricing, making complex document analysis affordable.
Discover how organizations across industries are leveraging GPT-4.1 to solve complex challenges.
GPT-4.1 transforms the development workflow with its enhanced coding capabilities.
The extended context window enables unprecedented analysis of lengthy legal documents.
GPT-4.1 excels at analyzing complex financial data and reports with high accuracy.
Unlock new creative possibilities with GPT-4.1's improved instruction following and context awareness.
The 1M token context window revolutionizes research and academic workflows.
Enhanced context awareness enables GPT-4.1 to handle complex customer interactions with ease.
Lead Developer, TechForward
"GPT-4.1's code generation capabilities have transformed our development process. The accuracy of its suggestions and minimal extraneous edits have significantly improved our team's productivity."
Legal Analyst, Lawton & Partners
"The 1 million token context window is a game-changer for legal document analysis. We can now review entire contracts and agreements in one go, saving hours of manual work while improving accuracy."
Content Director, MediaMind
"GPT-4.1's improvement in following complex instructions has been remarkable. It maintains the requested style and tone consistently throughout long-form content, which has streamlined our editorial process."
Research Lead, MediScience Labs
"As researchers, we deal with vast amounts of scientific literature. GPT-4.1's ability to process and synthesize information from multiple papers has cut our literature review time in half."
Multiple ways to experience the power of GPT-4.1 without a paid subscription.
OpenAI offers limited access to GPT-4.1 through the free tier of ChatGPT.
Developers can access GPT-4.1 through OpenAI's API with free credits for new accounts.
Students and educators can access GPT-4.1 through special educational programs.
Make the most of limited interactions by creating clear, precise prompts. Be specific about your requirements and the format you need.
When working within context limitations, focus on the most relevant information. Summarize background details and be concise in your requests.
Use the free tier during off-peak hours to experience faster response times and fewer rate limitations.
Take advantage of GPT-4.1's enhanced coding capabilities, which show the most significant improvements over previous models.
Divide larger projects into smaller, manageable chunks that work within the free tier's limitations.
| Feature | Free Access | Paid Access |
|---|---|---|
| Usage Limits | Limited | Unlimited |
| Response Speed | Standard | Priority |
| Context Window | Restricted | Full 1M tokens |
| Advanced Features | Basic | Complete |
| Peak-time Access | May queue | Guaranteed |
Upgrade to a paid plan for unlimited access to all features and capabilities.
Get answers to common questions about GPT-4.1.
GPT-4.1 is OpenAI's latest large language model released on April 14, 2025. It features significantly improved coding capabilities (54.6% on SWE-bench Verified), better instruction following (87.4% on IFEval), and an extended context window of up to 1 million tokens. Compared to GPT-4o, it produces more precise code edits, follows complex instructions more accurately, and can process much longer documents in a single prompt.
GPT-4.1 is available in three pricing tiers: GPT-4.1 at $2.00 per million input tokens and $8.00 per million output tokens, GPT-4.1 mini at $0.40 per million input tokens and $1.60 per million output tokens, and GPT-4.1 nano at $0.10 per million input tokens and $0.40 per million output tokens. OpenAI also offers a 75% discount for cached prompts, making repeated queries more cost-effective. Limited free access is available through ChatGPT's free tier, new API account credits, and educational programs.
The 1 million token context window enables processing of extremely lengthy documents and conversations in a single prompt. This is particularly valuable for:
There are several ways to access GPT-4.1 for free:
Free access typically has limitations on usage quotas, context window size, and may experience rate limits during peak times.
GPT-4.1 excels in various domains, including:
Its improved coding capabilities, instruction following, and extended context window make it particularly valuable for tasks requiring precision and handling large amounts of information.
GPT-4.1 represents a significant advancement in AI capabilities compared to both previous OpenAI models and competitors. Key comparisons include:
While other AI models from competitors may excel in specific niches, GPT-4.1's comprehensive improvements across coding, instruction following, and context handling make it a versatile and powerful general-purpose AI model.