deepseek-v3.2 vs kimi-k2.6

Capability for the money

Not the same tier: kimi-k2.6 leads deepseek-v3.2 by 27.0 points on following instructions (IFBench), for 5.5× the price.

1. Following instructions

   extraction, classification, sticking to a format

   IFBench
   deepseek-v3.249.0%

   kimi-k2.676.0%

   kimi-k2.6 +27.0

2. Tool use

   multi-step agent work against real tools

   τ²-bench
   deepseek-v3.278.9%

   kimi-k2.695.9%

   kimi-k2.6 +17.0

3. Hard reasoning

   graduate-level science and multi-step analysis

   GPQA Diamond
   deepseek-v3.275.1%

   kimi-k2.691.1%

   kimi-k2.6 +16.0

Averaged over the benchmarks both models sat, deepseek-v3.2 returns 218.3 points of pass rate per dollar of blended price against 51.2 — the ratio that decides whether the dearer model earns its rate.

Each figure is a pass rate: the share of that test set the model answered correctly, measured by a third party rather than self-reported by the lab. Rows are different tests of different difficulty, so compare a model against the other model, not one row against the next. No published coding score for one of these models, so that lens is omitted. And a benchmark is not your workload — it narrows the shortlist, it does not pick for you.