Processing or generating 100,000 words (approximately 133,300 tokens) with Grok 2 Vision ranges from $0.0667 (cached input) to $1.333 (full generation).
You send 100,000 words as prompt context, documentation, or background knowledge.
Grok 2 Vision drafts a complete 100,000 words article, chapter, or code module.
80,000 words input prompt + 20,000 words output reply.
| Model | Provider | Input Cost (100,000 words) | Cached Input Cost | Output Cost (100,000 words) | Context Limit |
|---|---|---|---|---|---|
| Grok 2 Vision (Current) | xai | $0.2666 | $0.0667 | $1.333 | 128,000 |
| GPT-5.6 Sol | openai | $0.64 | $0.064 | $3.84 | 1,050,000 |
| GPT-5.6 Cyber | openai | $1.60 | $0.16 | $9.60 | 1,050,000 |
| GPT-5.5 Standard | openai | $0.64 | $0.064 | $3.84 | 512,000 |
| GPT-5.5 Pro | openai | $3.84 | $0.384 | $23.04 | 512,000 |
| o3-pro (Frontier Reasoning) | openai | $2.56 | $0.256 | $10.24 | 1,000,000 |
For Grok 2 Vision (tiktoken_cl100k tokenizer), 100,000 words is approximately 133,300 tokens (an average ratio of 1.33 tokens per word in English). Code, technical vocabulary, and non-English scripts will have higher token densities.
Sending 100,000 words as input costs $0.2666 (or $0.0667 with prompt caching). Generating 100,000 words as output costs $1.333. 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 $0.0667). For non-urgent asynchronous batch processing, 24-hour batch queues provide a flat 50% discount across all tokens (costing $0.1333 for input, $0.6665 for output).
At an average human reading speed of 250 words per minute, 100,000 words takes approximately 400 minutes to read. In contrast, Grok 2 Vision can process or generate this text in seconds.