Extraction
Clinical Note SOAP Structuring
Given synthetic clinician dictation, convert stated information into subjective, objective, assessment, and plan sections without adding diagnoses or facts.
Healthcare & Life SciencesPer clinical note
Not benchmarked
No result on this tier
Token profile pending
— /100K
Run this tier to add it
—
Gemma 4 E4B IT
E4B · Off-the-shelf
324 in · 321 out · 5.36s p50
$4.07 /100K
Current price · TrustedRouter
View token rates · observed 9/11/2026
TrustedRouter · $0.0211 input · $0.1055 output /1M
63%
Cheapest
DeepSeek V3
Hosted · Off-the-shelf
303 in · 365 out · 4.01s p50
$44.50 /100K
Best current price · TrustedRouter
Compare 2 router rates · observed 9/11/2026
TrustedRouter · $0.3376 input · $0.9390 output /1M
OpenRouter · $0.2900 input · $1.14 output /1M
55%
GPT-5.4
Hosted · Off-the-shelf
306 in · 469 out · 3.81s p50
$780 /100K
Best current price · OpenRouter
Compare 2 router rates · observed 9/11/2026
OpenRouter · $2.50 input · $15.00 output /1M
TrustedRouter · $2.64 input · $15.82 output /1M
65%
All 12 model results and methodology
Generated 100/100 examples
| Model | Tier | Quality | Judged | Scenario cost |
|---|---|---|---|---|
| Nemotron Nano 9B v2 | small | 3% | 100/100 | $27.38 |
| Qwen3 235B A22B | mid | 23% | 100/100 | $87.04 |
| DeepSeek V3 | mid | 55% | 100/100 | $44.50 |
| Mistral Large 2407 | mid | 34% | 100/100 | $82.24 |
| Arcee Trinity Large Thinking | mid | 32% | 100/100 | $91.40 |
| GPT-5.4 | frontier | 65% | 100/100 | $780 |
| Claude Opus 4.7 | frontier | 59% | 100/100 | $1642 |
| Gemini 3.1 Pro Preview | frontier | 37% | 100/100 | $2191 |
| Gemma 4 E4B IT | small | 63% | 100/100 | $4.07 |
| Granite 4.1 8B | small | 29% | 100/100 | $5.17 |
| Ministral 8B Instruct 2410 | small | 31% | 100/100 | $10.15 |
| Qwen3 4B Instruct 2507 | small | 5% | 100/100 | $20.72 |
LLM-judge pass rate on 100 synthetic examples. Generator: gpt-5.2. Judge: gpt-5.2. Evaluated 2026-09-09T07:12:02.797Z.
Directional: measured on a synthetic eval set generated by drydock. Cost/latency are not yet captured for taskrouter-run benchmarks.
drydock