llama-3.1-70b-instruct vs llama-3.1-8b-instant

Capability for the money

llama-3.1-70b-instruct is 17.4× dearer for 15.0 points on hard reasoning (GPQA Diamond) — worth it only if that is your bottleneck.

1. Following instructions

   extraction, classification, sticking to a format

   IFBench
   llama-3.1-70b-instruct34.4%

   llama-3.1-8b-instant28.6%

   llama-3.1-70b-instruct +5.8

2. Coding

   writing and fixing code

   LiveCodeBench
   llama-3.1-70b-instruct23.2%

   llama-3.1-8b-instant11.6%

   llama-3.1-70b-instruct +11.6

3. Tool use

   multi-step agent work against real tools

   τ²-bench
   llama-3.1-70b-instruct15.2%

   llama-3.1-8b-instant16.4%

   llama-3.1-8b-instant +1.2

4. Hard reasoning

   graduate-level science and multi-step analysis

   GPQA Diamond
   llama-3.1-70b-instruct40.9%

   llama-3.1-8b-instant25.9%

   llama-3.1-70b-instruct +15.0

Averaged over the benchmarks both models sat, llama-3.1-8b-instant returns 358.7 points of pass rate per dollar of blended price against 28.4 — 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. And a benchmark is not your workload — it narrows the shortlist, it does not pick for you.