Best · Best AI model under 32B for factuality — October 2026

Best AI model under 32B for factuality — October 2026

As of Oct 14, 2026, Gemma 4 31B & Gemma 4 26B A4b & Qwen3 8 27B are tied for state of the art in Factuality, models at or under 32B (since Oct 8, 2026; sources LMArena Text factuality; methodology v1.0).

Category board for Factuality

Top 10

Ranks

  1. #1 Gemma 4 31B
  2. #1 Gemma 4 26B A4b
  3. #1 Qwen3 8 27B
  4. #2 Qwen3 5 27B
  5. #3 Granite 4.2 30B
  6. #4 Nvidia Nemotron 3.5 Lightning 30B A3b Nvfp4
  7. #5 Nvidia Nemotron 3 Nano 30B A3b Bf16
  8. #6 Granite 4.2 8B
  9. #7 Olmo 3.1 32B Instruct
  10. #8 Granite 4.2 3B
Recent crowns

Last crown changes

  • Oct 8, 2026 — Factuality: Gemma 4 31B and Gemma 4 26B A4b and Qwen3 8 27B are co-SOTA
Weights

Which benchmarks count

Who is SOTA in Factuality, models at or under 32B?

As of Oct 14, 2026, Gemma 4 31B & Gemma 4 26B A4b & Qwen3 8 27B are tied for state of the art in Factuality, models at or under 32B (since Oct 8, 2026; sources LMArena Text factuality; methodology v1.0).

Since when is the Factuality, models at or under 32B crown current?

The current crown in Factuality, models at or under 32B has held since Oct 8, 2026.