gpt-5.6 vs gpt-5.6-terra

Capability for the money

gpt-5.6 costs 2.5× more than gpt-5.6-terra without measurably beating it on any benchmark they share — on that evidence the extra spend buys nothing you can point at.

1. Following instructions

   extraction, classification, sticking to a format

   IFBench
   gpt-5.672.7%

   gpt-5.6-terra71.2%

   gpt-5.6 +1.5

2. Tool use

   multi-step agent work against real tools

   τ²-bench
   gpt-5.685.1%

   gpt-5.6-terra86.3%

   gpt-5.6-terra +1.2

3. Hard reasoning

   graduate-level science and multi-step analysis

   GPQA Diamond
   gpt-5.694.1%

   gpt-5.6-terra92.5%

   gpt-5.6 +1.6

Averaged over the benchmarks both models sat, gpt-5.6-terra returns 18.5 points of pass rate per dollar of blended price against 7.5 — 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.