Falcon 2 vs Llama 3.1 8B

Falcon 2 11B vs Llama 3.1 8B: A Moving Comparison

Meta released Llama 3.1 8B after TII launched Falcon 2 11B, with a longer context window and a larger training set. That makes the original Falcon 2 comparison a snapshot of an earlier generation.

Back to the overview

What changed with Llama 3.1

  • A much larger context window than the 8,192 tokens configured on Falcon 2 11B.
  • A larger pretraining corpus than Llama 3 8B.
  • Tool-calling support in the instruction-tuned variants.

Why the old headline no longer applies

TII benchmarked Falcon 2 11B against Llama 3 8B, not 3.1. Comparing a May 2024 model against a later release is a different question, and neither vendor published that head-to-head. Anyone claiming a clean winner is extrapolating.

What to compare instead

  • Context: 8,192 tokens on Falcon 2 11B is a hard configuration limit for long-document work.
  • Vision: Llama 3.1 8B is text only; Falcon 2 offers a matching VLM.
  • Licence and acceptable use terms, which differ between the two.
  • Your own task accuracy, which is the only number that survives a leaderboard refresh.

Related pages

Falcon 2 vs Llama 3 8BContext lengthFalcon 2 benchmarks

Sources

falcon2.lol is an independent third-party site. It is not affiliated with, endorsed by, sponsored by or operated by the Technology Innovation Institute, and it does not speak for TII. Every number on this page is attributed to the source listed above.

Falcon 2 11B vs Llama 3.1 8B: A Moving Comparison