Settings

Settings

Application configuration and preferences

AI Model

Select the AI model used for tender analysis and solution auto-fill.

Claude Sonnet 4.6Balanced
Input£2.37 / MTokCache write£2.96 / MTokCache read£0.24 / MTokOutput£11.85 / MTok

Best balance of speed and intelligence for most RFP tasks.

Cost per Tender

Estimated AI cost to process one tender through the full RFP Edge pipeline.

200 pg
15
Claude Sonnet 4.6£1.12
Claude Opus 4.6£5.58
Claude Haiku 4.5£0.30

Pipeline breakdown — Claude Sonnet 4.6

Document analysis
132k tok£0.41
Triage scoring
2.0M tok£0.52
Solution auto-fill
138k tok£0.08
USP matching
134k tok£0.05
Executive summary
138k tok£0.06
Cache write (1st call)130k tok — £0.39
Cache reads (18 calls)2.3M tok — £0.55
Non-cached input15k tok — £0.03
Output12k tok — £0.14
Total (2.5M tokens)£1.12

Estimates based on ~650 tokens/page using Files API + Prompt Caching. The document is uploaded once and cached on the first call — subsequent calls read from cache at ~10% of input cost. Prices converted at estimated rate: $1 = £0.7900.