A standard 25 pages document (~12,500 words / 16,663 tokens) costs $0.000833 to ingest and $0.000873 to generate a comprehensive executive summary using Llama 3.1 8B Instruct.
Feed the entire 25 pages document into context to extract fields, entities, or answer queries.
Ingest 25 pages and generate a structured 500-token executive summary with key takeaways.
Translate or reformat the entire 25 pages document into an equivalent length output.
| Model | Provider | Ingestion (25 pages) | Cached Ingestion | Summary Cost | Context Limit |
|---|---|---|---|---|---|
| Llama 3.1 8B Instruct (Current) | meta | $0.000833 | $0.000208 | $0.000873 | 128,000 |
| GPT-5.6 Luna | openai | $0.0032 | $0.00032 | $0.0038 | 1,050,000 |
| GPT-5.4 mini | openai | $0.012 | $0.0012 | $0.0143 | 256,000 |
| GPT-5.4 nano | openai | $0.0032 | $0.00032 | $0.003825 | 128,000 |
| GPT-4o mini | openai | $0.0024 | $0.0012 | $0.0027 | 128,000 |
| Claude Haiku 4.5 | anthropic | $0.0165 | $0.00165 | $0.019 | 1,000,000 |
Assuming a standard single-spaced document with ~500 words per page, a 25 pages document contains approximately 12,500 words, which translates to roughly 16,663 tokens using Llama 3.1 8B Instruct's tokenizer. Dense PDF tables, legal boilerplate, and OCR scans may increase this by 20–30%.
Llama 3.1 8B Instruct has a context window of 128,000 tokens. A 25 pages document consumes only 13.02% of its total window, easily fitting in a single prompt without requiring chunking or vector search.
Ingesting the document and outputting a concise 500-token executive summary costs $0.000873. Using 24-hour async batch API queues drops this cost to $0.000437.
If your workflow repeatedly queries or chats with this same 25 pages document, prompt caching reduces subsequent turn input costs by up to 90% (to $0.000208 per turn instead of $0.000833).