← Back to changelog

April 16, 2025

OpenAI o3 and o4-mini integration for playground, evaluations and cost tracking

Marlies Mayerhofer

Langfuse launches same day full compatibility with OpenAI's latest models o3 and o4-mini

Some highlights about o3 and o4-mini from the OpenAI release notes:

• Improved Instruction Following: Both models demonstrate enhanced instruction following and provide more useful, verifiable responses than their predecessors
• More Natural Interaction: Responses feel more conversational and personalized by leveraging memory and past conversation context

As OpenAIs best reasoning model yet, o3:

• Advanced Reasoning: o3 is OpenAI’s most powerful reasoning model, setting new state-of-the-art benchmarks across coding, math, science, and visual perception
• Reduced Error Rate: Makes 20 percent fewer major errors than OpenAI o1 on difficult real-world tasks, especially in programming, business consulting, and creative ideation

o4-mini for cost-effective reasoning:

• Cost-Effective Performance: o4-mini achieves remarkable performance for its size and cost, particularly excelling in math, coding, and visual tasks
• Higher Usage Capacity: Supports significantly higher usage limits than o3, making it ideal for high-volume applications that benefit from reasoning

Learn more

• Langfuse LLM playground
• Langfuse model usage and cost tracking