Accuracy-cost value frontier: at $0.10 per million tokens, CDPK performance rose from 50% (April 2024) to 82% (June 2025)¶
Claim · All claims
Evidence · 1 study · designr2·q2
Currency · newest study 2025 · not yet checked against its sources · not reviewed by a person · recheck by 2028: tool and model accuracy changes quickly
Subclaims¶
Reading the evidence codes
q is the design tier, i the impact, n the sample, and the kind with its r
the rigour judged within that kind: what each code means.
This claim: no standardized effect is recorded in observations/ for it yet, so no benchmark reading is possible.
q2 i? The Pareto value frontier of CDPK accuracy versus inference cost has improved rapidly over 18 months, with performance at $0.10 per million tokens increasing from 50% to 70% to 82% between April 2024 and June 2025. → Lelièvre 2025
Evidence¶
Lelièvre 2025¶
Lelièvre, M., Waldock, A., Liu, M., Valdes Aspillaga, N., Mackintosh, A., Ogando Portela, M. J., Lee, J., Atherton, P., Ince, R. A. A., & Garrod, O. G. B. (2025). Benchmarking the Pedagogical Knowledge of Large Language Models. https://arxiv.org/abs/2506.18710 doi:10.48550/arxiv.2506.18710
q2 · i? · design · r2
Numerical analysis of benchmarked models' input-token costs against CDPK accuracy, charting the Pareto value frontier (convex upper hull) and its evolution over time, shown in Figure 8. The article reports the $0.10/Mtoken performance trajectory of 50% to 70% to 82%.
"For example, at $0.10 per million tokens, the performance has increased from 50% (April 2024) to 70% (November 2024) to 82% (June 2025)."
Discussion¶
Related Claims¶
- The benchmark procedure is stable under stochastic decoding, with a maximum accuracy standard deviation of 0.43% over 20 repeated runs — related
- CDPK performance scales with model size, with a sharp Pareto-frontier drop-off below around 8B parameters — related
- Estimated human baseline on the CDPK benchmark is approximately 50%, based on aggregate Chilean teacher exam results — related
- May 2026 token telemetry was cache-dominant: 82.9% of 73,950,305 recorded tokens were cache reads, suggesting persistent workflows shift the economic unit from cost per token to cost per completed artifact
- LLM accuracy on the CDPK pedagogy benchmark ranges from 28% to 89%, with closed-weight reasoning models dominating the top of the leaderboard — related
- Distillation increases misclassification of ambiguous To Some Extent tutor responses as Yes — related
Page metadata
| Field | Value |
|---|---|
type |
claim |
id |
cdpk-value-frontier-rapid-cost-performance-improvement |
title |
Accuracy-cost value frontier: at $0.10 per million tokens, CDPK performance rose from 50% (April 2024) to 82% (June 2025) |
description |
Accuracy-cost value frontier: at $0.10 per million tokens, CDPK performance rose from 50% (April 2024) to 82% (June 2025) |
status |
review |
evidence_strength |
moderate |
newest_evidence |
2025 |
recheck_by |
2028 |
generated |
by process:wiki-ingest on 2026-10-10 |
trust |
unverified — no one has confirmed this page's content |
sources |
1 — lelièvre-2025 |
Frontmatter is the page's machine-readable record — see the Schema & Guide.