v0.2 · Model comparison

Compare models across every published benchmark.

Choose a comparison priority, inspect the same models across all nine published benchmarks, and connect every score and resource total to its recorded evidence.

Choose a comparison

Automatic priorities consider only models with all nine published benchmark receipts.

Kimi K2.6

9 of 9 run receipts

Kimi K3

9 of 9 run receipts

Claude Opus 5

9 of 9 run receipts

Benchmark scores

Published normalized scores on a 0–100 scale. Every exact value opens its run receipt.

BenchmarkKimi K2.6Kimi K3Claude Opus 5
Accounting
Accounting Knowledge94.655691.744492.3556
HGB Reporting90.177888.433386.5000
IFRS Reporting91.111191.444488.1667
Financial Analysis87.542766.500079.4111
Accounting Capstone91.893388.667689.7238
Corporate Finance
Capital Budgeting65.000086.666780.0000
Cost of Capital & Financing93.333391.666786.6667
Valuation & Financial Modeling88.333395.000096.6667
M&A, LBO & Capital Allocation98.3333100.000096.6667

Score profile

Each point is one observed benchmark score. Exact published values are listed in the score matrix.

Normalized benchmark score profile3 selected models across 9 benchmarks. See the adjacent score matrix for exact values.1008060AccountingHGBIFRSFSAAcc. CapCapBudWACCValuationM&A/LBOKimi K2.6 benchmark score seriesKimi K3 benchmark score seriesClaude Opus 5 benchmark score series
  1. Kimi K2.6
  2. Kimi K3
  3. Claude Opus 5

Resource evidence

Aggregated only from the same nine linked run receipts.

MeasureKimi K2.6Kimi K3Claude Opus 5
Mean score88.988.988.5
Total recorded cost$13.35$11.75$15.12
Total tokens3,599,4031,504,5341,734,772