model-savings

1# Model Savings
2 
3Quantify how much spend you would recover by moving eligible work to a cheaper model.
4 
5## Input
6 
7The user provides: **$ARGUMENTS**
8 
9This is the routing question — e.g. `"Opus → Sonnet"`, `"move simple work to Haiku"`,
10or empty (analyze every premium model against the next tier down). If no target family
11is named, default to proposing the next-cheaper tier per model and say so.
12 
13## Data Sources
14 
15| Endpoint | Returns |
16|----------|---------|
17| `GET /api/pricing` | `{ pricing: [{ model_pattern, display_name, input_per_mtok, output_per_mtok, cache_read_per_mtok, cache_write_per_mtok }] }` — the rate card for every family |
18| `GET /api/pricing/cost` | `{ total_cost, breakdown: [{ model, input_tokens, output_tokens, cache_read_tokens, cache_write_tokens, cost, matched_rule }] }` — current spend and the exact token mix per model |
19| `GET /api/sessions?limit=200` | Sessions with `model`, inline `cost`, and `metadata` (turn_count, thinking_blocks) — used to judge which work is *eligible* to downshift |
20| `GET /api/analytics` | `agent_types`, `tool_usage`, `total_subagents` — corroborate which task types are low-complexity and safe to route cheaper |
21 
22## Savings method
23 
24For each candidate model in the cost `breakdown`, re-price its **exact token mix** at the target family's rates:
25 
26```
27cost_at_target = (input_tokens      / 1M) × target.input_per_mtok
28               + (output_tokens     / 1M) × target.output_per_mtok
29…