AI Potluck
Model components / Fine-tuned / chat models

VibeThinker-3B

WeiboAI

WeiboAI's VibeThinker-3B, an MIT-licensed 3B reasoning model fine-tuned from Qwen2.5-3B / Qwen2.5-Coder-3B and specialized for verifiable math, coding and STEM. Claims frontier-range scores on narrow reasoning benchmarks despite its size; not intended for tool-calling or agentic use.

MIT-licensed weights, but a fine-tune of the open-weight Qwen2.5 line without a fully open data/training pipeline -> open_weights, not open_source. Narrow specialist.

Openness

3 medium confidence
3.0
weights
open(MIT)
base
open_weights(Qwen2.5)
data
partial(RL/SFT data not fully released)
license
MIT

MIT-licensed weights, but a fine-tune of the open-weight Qwen2.5 line without a fully open/reproducible data+training pipeline -> open_weights tier, not open_source.

Adoption

2 low confidence
2.0

Recently released specialist model with early attention for its size; directional estimate (not a measured usage figure).

Capability

3 medium confidence
3.0

Frontier-range on narrow verifiable math/code reasoning despite 3B, but explicitly not for general/agentic use -> mid-tier on a general-capability scale.

Unchanged since 2026-06-30 (last edited, not re-checked)