Best LLMs for content-summarization autogenerated — DTP Benchmark
Models
Frontier on this task: GPT-5.4 nano at 8.92 / 10. Quality bar at 95%: 8.47.
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 |
|---|---|---|---|---|---|
| GPT-5.4 nano best OpenAI | 8.92 / 10 CI [8.79, 9.04] | n=100 · ranked | $0.000954 | (anchor) | batch |
| Kimi K2.6 Moonshot AI | 8.61 / 10 CI [8.45, 8.78] | n=100 · ranked | $0.003974 | — | batch |
Typical call shape for this task: 1666 input tokens → 1260 output tokens, EMA-tracked from production traffic. Blended cost = (in × in_price + out × out_price), rounded to 6 decimals.