OpenAI provides industry-leading AI models:
- GPT-5 Series - Latest generation language models
- o-series - Advanced reasoning capabilities
- Multimodal - Vision and audio understanding
- Codex - Specialized code generation
Available Models
GPT-5.1 Series
Enhanced GPT-5 models with improved performance.GPT-5 Series
Latest generation models with enhanced capabilities.GPT-4o Series
Multimodal models with vision capabilities.GPT-4.1 Series
Reliable, proven models for production.o-series (Reasoning Models)
Advanced reasoning models that “think” before responding.Best Practices
Model Selection
Use GPT-5 for
- Latest capabilities
- Best overall performance
- Production applications
- Cutting-edge features
Use GPT-4o for
- Vision tasks
- Multimodal needs
- Consistent quality
- Stable production
Use o-series for
- Complex reasoning
- Math problems
- Code debugging
- Research tasks
Use Mini/Nano for
- High-volume requests
- Simple tasks
- Cost optimization
- Fast responses
Prompt Caching
Optimize costs with prompt caching:
- First request: Full cost ($1.25/M for GPT-5)
- Cached requests: 90% discount ($0.13/M)
- Cache duration: 5 minutes of inactivity
- Chatbots with large system prompts
- Document analysis with consistent context
- Multi-turn conversations
- RAG applications with static knowledge
Context Windows
OpenAI models support large context:- GPT-4o: Up to 128K tokens (~100K words)
- GPT-5: Up to 128K tokens (~100K words)
- o-series: Up to 128K tokens (includes reasoning tokens)
Support
Need help with OpenAI integration?OpenAI Documentation
Official OpenAI documentation
OpenAI Pricing
Official pricing details
Splox Docs
Browse our guides
Community
Get help from the community