$7.5553
Total spend (30d)
8,426 calls
$7.56
Projected monthly (run-rate)
720h window
$0.8967
Cost / 1,000 answers
8,426 requests
11.2s
Latency p95
avg 4.68s
Cost by model
| Model | Calls | p95 latency | avg prompt | Cost |
|---|---|---|---|---|
| google/gemini-2.5-flash-lite | 2853 | 1.0s | 451 tok | $0.1839 |
| openai/gpt-4o-mini | 2663 | 3.3s | 203 tok | $0.4487 |
| google/gemini-2.5-flash | 789 | 21.2s | 2141 tok | $1.7740 |
| google/gemini-3.1-flash-lite | 739 | 11.4s | 12908 tok | $4.5579 |
| deepseek/deepseek-v4-pro | 652 | 3.0s | 42 tok | $0.0898 |
| deepseek/deepseek-v4-flash | 215 | 312.7s | 13067 tok | $0.4769 |
| gpt-4o-mini | 204 | 2.3s | 64 tok | $0.0062 |
| typesafe/jev-latest | 199 | 0.4s | 731 tok | $0.0061 |
| deepseek/deepseek-chat | 51 | 7.6s | 82 tok | $0.0015 |
| deepseek/deepseek-v4-flash-0731 | 20 | 38.1s | 504 tok | $0.0026 |
Cost by feature
| App | Feature | Calls | Spend | Cost/answer |
|---|---|---|---|---|
| droitassist | search_answer_gen | 987 | $6.2184 | $0.0063 |
| droitassist | smoke_test_monitor | 159 | $0.5236 | $0.0033 |
| sowknow | unknown | 2162 | $0.3665 | $0.0002 |
| droitassist | unattributed | 1330 | $0.1091 | $0.0001 |
| droitassist | search_guard | 1690 | $0.1082 | $0.0001 |
| droitassist | grounding_expansion | 1102 | $0.0692 | $0.0001 |
| sowknow | search_agent | 347 | $0.0592 | $0.0002 |
| droitassist | unknown | 24 | $0.0389 | $0.0016 |
| sowknow | collection_search | 158 | $0.0292 | $0.0002 |
| sowknow | copilot | 11 | $0.0138 | $0.0013 |
| kodam | copilot | 196 | $0.0061 | $0.0000 |
| droitassist | rag_llm | 10 | $0.0057 | $0.0006 |
| sowknow | intent | 149 | $0.0054 | $0.0000 |
| sowknow | unattributed | 8 | $0.0008 | $0.0001 |
| sowknow | legal_search | 9 | $0.0004 | $0.0000 |
| sowknow | interactive_chat | 16 | $0.0002 | $0.0000 |
| kodam | demand_forecast | 3 | $0.0001 | $0.0000 |
| kodam | quantite_stocker | 3 | $0.0001 | $0.0000 |
| kodam | unknown | 4 | $0.0000 | $0.0000 |
| droitassist | le_magistrat | 7 | $0.0000 | $0.0000 |
Token incinerators (context bloat)
| Feature | Model | Reqs | avg prompt | Cost |
|---|---|---|---|---|
| smoke_test_monitor | deepseek/deepseek-v4-flash | 8 | 14257 tok | $0.0177 |
| unknown | deepseek/deepseek-v4-flash | 2 | 13542 tok | $0.0042 |
| smoke_test_monitor | google/gemini-3.1-flash-lite | 62 | 13538 tok | $0.3950 |
| search_answer_gen | deepseek/deepseek-v4-flash | 197 | 13480 tok | $0.4531 |
| search_answer_gen | google/gemini-3.1-flash-lite | 662 | 12988 tok | $4.1250 |
| unattributed | deepseek/deepseek-v4-flash | 1 | 12738 tok | $0.0018 |
| search_answer_gen | google/gemini-2.5-flash | 122 | 12563 tok | $1.6404 |
| unattributed | google/gemini-3.1-flash-lite | 2 | 12272 tok | $0.0120 |
Live span feed
| Time (UTC) | App | Feature | Model | Status | Tokens | Cost |
|---|---|---|---|---|---|---|
| 2026-09-28 14:23:57 | droitassist | search_guard | typesafe/jev-latest | error | 0→0 | $0.0000 |
| 2026-09-28 14:23:52 | droitassist | search_guard | typesafe/jev-latest | ok | 1275→358 | $0.0001 |
| 2026-09-28 14:23:52 | droitassist | search_guard | google/gemini-2.5-flash-lite | ok | 462→49 | $0.0001 |
| 2026-09-28 14:23:42 | droitassist | search_guard | google/gemini-2.5-flash-lite | ok | 455→34 | $0.0001 |
| 2026-09-28 14:23:25 | droitassist | search_guard | typesafe/jev-latest | ok | 1268→361 | $0.0001 |
| 2026-09-28 14:23:25 | droitassist | search_guard | google/gemini-2.5-flash-lite | ok | 456→34 | $0.0001 |
| 2026-09-28 14:23:22 | droitassist | search_answer_gen | google/gemini-3.1-flash-lite | ok | 10873→2081 | $0.0058 |
| 2026-09-28 14:22:59 | droitassist | grounding_expansion | google/gemini-2.5-flash-lite | ok | 406→49 | $0.0001 |
| 2026-09-28 14:22:44 | droitassist | search_answer_gen | google/gemini-3.1-flash-lite | ok | 13899→1994 | $0.0065 |
| 2026-09-28 14:22:23 | droitassist | grounding_expansion | google/gemini-2.5-flash-lite | ok | 399→53 | $0.0001 |
| 2026-09-28 14:22:13 | droitassist | search_guard | typesafe/jev-latest | ok | 1279→361 | $0.0001 |
| 2026-09-28 14:22:12 | droitassist | search_guard | google/gemini-2.5-flash-lite | ok | 466→48 | $0.0001 |
| 2026-09-28 14:22:07 | droitassist | search_guard | typesafe/jev-latest | ok | 1276→360 | $0.0001 |
| 2026-09-28 14:22:07 | droitassist | search_guard | google/gemini-2.5-flash-lite | ok | 466→67 | $0.0001 |
| 2026-09-28 13:24:48 | droitassist | search_guard | typesafe/jev-latest | ok | 1278→362 | $0.0001 |
| 2026-09-28 13:24:47 | droitassist | search_guard | google/gemini-2.5-flash-lite | ok | 465→52 | $0.0001 |
| 2026-09-28 13:24:39 | droitassist | search_guard | typesafe/jev-latest | ok | 1280→358 | $0.0001 |
| 2026-09-28 13:24:38 | droitassist | search_guard | google/gemini-2.5-flash-lite | ok | 463→30 | $0.0001 |
| 2026-09-28 13:24:34 | droitassist | search_guard | typesafe/jev-latest | ok | 1280→359 | $0.0001 |
| 2026-09-28 13:24:34 | droitassist | search_guard | google/gemini-2.5-flash-lite | ok | 468→31 | $0.0001 |
p95 latency max bar = 312.7s · window 30d