Processing or generating 500,000 words (approximately 640,000 tokens) with OpenRouter Auto-Best Router ranges from $0.16 (cached input) to $1.92 (full generation).
You send 500,000 words as prompt context, documentation, or background knowledge.
OpenRouter Auto-Best Router drafts a complete 500,000 words article, chapter, or code module.
400,000 words input prompt + 100,000 words output reply.
| Model | Provider | Input Cost (500,000 words) | Cached Input Cost | Output Cost (500,000 words) | Context Limit |
|---|---|---|---|---|---|
| OpenRouter Auto-Best Router (Current) | openrouter | $0.64 | $0.16 | $1.92 | 1,000,000 |
| GPT-5.6 Terra | openai | $1.28 | $0.128 | $7.68 | 1,050,000 |
| GPT-5.4 Workhorse | openai | $1.60 | $0.16 | $9.60 | 256,000 |
| o3-mini | openai | $0.704 | $0.352 | $2.816 | 200,000 |
| o4-mini | openai | $0.704 | $0.176 | $2.816 | 256,000 |
| o1-mini | openai | $0.704 | $0.352 | $2.816 | 128,000 |
For OpenRouter Auto-Best Router (o200k_base tokenizer), 500,000 words is approximately 640,000 tokens (an average ratio of 1.28 tokens per word in English). Code, technical vocabulary, and non-English scripts will have higher token densities.
Sending 500,000 words as input costs $0.64 (or $0.16 with prompt caching). Generating 500,000 words as output costs $1.92. Output tokens are more expensive because autoregressive token generation requires significantly more computation per token.
Yes! If your input text is part of a repeated context, prompt caching saves 50% to 90% (costing $0.16). For non-urgent asynchronous batch processing, 24-hour batch queues provide a flat 50% discount across all tokens (costing $0.32 for input, $0.96 for output).
At an average human reading speed of 250 words per minute, 500,000 words takes approximately 2000 minutes to read. In contrast, OpenRouter Auto-Best Router can process or generate this text in seconds.