AI Model Pricing

Compare per-million-token pricing across 14 AI models. Find the best value for your budget.

ModelProviderInput / 1MOutput / 1MContextCoding
DeepSeek V4 Flashdeepseek$0.14/M$0.28/M262,14482
Gemini 2.5 Flashgoogle$0.30/M$2.50/M1,048,57676
Llama 4 Scoutmeta$0.30/M$1.00/M262,14474
GPT-5.4 miniopenai$0.75/M$4.50/M262,14476
Gemini 2.5 Progoogle$1.25/M$10.00/M1,000,00090
Grok 4.3xai$1.25/M$2.50/M262,14484
Mistral Medium 3.5mistral$1.50/M$7.50/M262,14482
Gemini 3.5 Flashgoogle$1.50/M$9.00/M1,048,57680
Llama 4 Maverickmeta$1.50/M$4.50/M262,14486
DeepSeek V4 Prodeepseek$1.74/M$3.48/M1,048,57693
GPT-5.4openai$2.50/M$15.00/M262,14492
Claude 4 Sonnetanthropic$3.00/M$15.00/M200,00094
Claude 4 Opusanthropic$5.00/M$25.00/M200,00096
GPT-5.5openai$5.00/M$30.00/M524,28897

Best Budget Models

Under $1/M input tokens — great for high-volume applications

Premium Models

$10+/M input tokens — maximum capability for demanding tasks

    AI Model Pricing Guide

    AI model pricing varies significantly across providers and model tiers. Most models charge per million tokens for both input (the text you send) and output (the text generated). Key factors affecting price include model size, capability level, and whether the model uses cached context. Budget models like GPT-4o mini and Gemini 2.0 Flash offer sub-$1/M input pricing, while premium frontier models like Claude 4 Opus command higher prices for maximum intelligence. Context window size also correlates with price — models with larger context windows (200K+, 1M) tend to cost more per token but can reduce the need for complex chunking strategies.