Pending benchmarksTracking since Oct 19, 2026
Is NABLA-FC state of the art?Pending benchmarks.Announced Oct 15, 2026; no independent results yet.
NABLAS's Japanese fact-checking model (Qwen3-8B fine-tune) plus a citing agent that surfaces primary sources with the verdict; UTokyo spin-off work from GENIAC phase 3.
Announced facts
The details so far
- Lab
- NABLAS Inc.
- Announced
- Oct 15, 2026
- API
- Not yet
- Open weights
- Published
- Parameters
- 8B (fine-tuned from Qwen3-8B)
- Type
- Japanese fact-checking LLM and citing agent
- Base
- Qwen/Qwen3-8B
- Languages
- Japanese and English
- License
- Apache-2.0
- Funding
- GENIAC phase 3 (lab announcement)
- Weights
- nablasinc/NABLA-FC on Hugging Face
Tags marked unsure are not stated on the lab's own pages.
Next steps
What would let us call it
Verdicts use only independent boards we already track. Two consecutive snapshots are required before any crown.
- Publicly announced with a source
- Usable by the public (API or weights)
- Listed on an independent board we track
- Enough results for a rank in a category
- Holds across two snapshots (crowns only)
Self-reported
What the lab claims
Self-reported. Shown for context; not used in any verdict.
Lab-reported manual eval (self-reported): NABLA-FC 32/33 on JSocialFact and 45/50 on FEVEROUS vs GPT-4o-mini and Gemini-3.1-lite-preview on their subset.
Receipts
Sources
LABNABLAS announcement on Xx.com/nablas_inc/status/2110537005163802693 · Oct 15, 2026LABNABLA-FC press releasenablas.com/post/nabla-fc · Oct 15, 2026REPONABLA-FC on Hugging Facehuggingface.co/nablasinc/NABLA-FC · Oct 15, 2026Tracking continues until an independent board lists it.