Best LLMs for claim_extraction autogenerated — DTP Benchmark
Models
Frontier on this task: DeepSeek V4 Pro at 7.45 / 10. Quality bar at 95%: 7.08.
point-estimate floor (CI low) · upper CI (less certain) · Bars sorted by blended cost; cheapest qualifier first. Greyed rows are MEDIUM+ models whose point estimate clears the bar but whose CI low does not.
Cost breakdown
| Model | Quality | Sample | Blended cost / call | Savings vs best | Mode |
|---|---|---|---|---|---|
| DeepSeek V4 Pro best DeepSeek | 7.45 / 10 CI [7.00, 7.91] | n=100 · high | $0.005749 | (anchor) | sync |
Typical call shape for this task: 2668 input tokens → 318 output tokens, EMA-tracked from production traffic. Blended cost = (in × in_price + out × out_price), rounded to 6 decimals.