gpt-4o-mini

By price, gpt-4o-mini ranks 15th cheapest of 68 on input and 15th on output. Its quality score of 80 ranks 48th of 68, and on value (quality score ÷ output price) it ranks 15th of 68. Its blended rate sits 91% below the catalog average.

In real billing terms: a support chatbot handling 50,000 turns a month (2,000 input / 500 output tokens each) costs about $30 per month on gpt-4o-mini, while a RAG workload of 20,000 queries (8,000 input / 800 output tokens each) runs about $34. Because output is priced 4.0× input here, long-generation tasks scale cost fastest — cap output length before you optimise anything else.

A 128K context window holds roughly 171 pages of A4 text (at ~750 tokens per page), ranking 53rd of 68 here. That covers typical RAG and multi-turn chat, though book-length inputs still need retrieval and chunking. Output runs at about 160 tok/s (16th fastest of 68), fast enough for real-time interaction and high-throughput batches.

We track 11 OpenAI models, and gpt-4o-mini's blended rate sits 93% below that provider's average. If your task tolerates a little less headroom, qwen3.5-flash (Alibaba) blends about 79% cheaper at a quality score of 82, and is worth benchmarking first. Going the other way, Hunyuan HY3 scores 87 on quality for about 62% less.

The closest model in OpenAI's line-up is gpt-5.6-luna ($0.20 input / $1.20 output, quality 90). Against it, gpt-4o-mini scores 10 points lower on quality, runs 10 tok/s faster (about 7%), has a 921K smaller context window, costs 71% less blended. When two models sit this close on price, the deciding factor is usually the quality ceiling — run both against your real prompts before picking.