# Blink v0.3 26B-A4B (NVFP4) vs torchcast-decision-27b (Torchcast AI, Qwen3.8-27B fine-tune)

As of Oct 15, 2026, Blink v0.3 26B-A4B (NVFP4) leads torchcast-decision-27b (Torchcast AI, Qwen3.8-27B fine-tune) in 0 of 1 shared categories.

Blink v0.3 26B-A4B (NVFP4): price per 1M tokens is unset in the catalog.

torchcast-decision-27b (Torchcast AI, Qwen3.8-27B fine-tune): price per 1M tokens is unset in the catalog.

Blink v0.3 26B-A4B (NVFP4): open weights.

torchcast-decision-27b (Torchcast AI, Qwen3.8-27B fine-tune): open weights.

- Which to pick, Decision models: As of Oct 15, 2026, torchcast-decision-27b (Torchcast AI, Qwen3.8-27B fine-tune) leads Blink v0.3 26B-A4B (NVFP4) in Decision models at #8 against #35.

As of Oct 15, 2026, Blink v0.3 26B-A4B (NVFP4) is ranked #35 in Decision models (since Oct 15, 2026; sources Decision Index Full; methodology v1.0).

As of Oct 15, 2026, torchcast-decision-27b (Torchcast AI, Qwen3.8-27B fine-tune) is ranked #8 in Decision models (since Oct 15, 2026; sources Decision Index Full and JevBench; methodology v1.0).