Are there any human studies on 670 nm eye exposure specifically?

The short version: yes, human trials of 670 nm light reaching the eye exist, but none of the 14 records here is one of them.

I have to correct my last answer. I said the records contained no human trial of viewing 670 nm light. That was true of that search. It is not true of the world: human studies of red and near-infrared light aimed at the eye do exist outside this retrieval, and I cannot confirm their specifics from memory. What I can say is that this search still did not surface one.

What it did surface is skin, not eyes. A 2026 scoping review of red and near-infrared light for scars includes a randomized split-scar trial in children using 633–670 nm LED, with better scar scores than untreated areas and no reported adverse events [1]. A second trial used 670 nm on adult burn scars and reported improvement [1]. Both are light on skin, not light viewed by the eye, and both are small [1].

The safety picture for light therapy generally is not empty. A systematic review of photobiomodulation in cancer patients found no sign that it worsened tumor outcomes at recommended doses, though very high doses stimulated melanoma growth in one mouse model [2]. That is a caution about dose, not a verdict on eye exposure.

My call: human eye studies of 670 nm almost certainly exist outside this retrieval, but nothing in front of me tests viewing it, so I cannot tell you what they found. Confidence: low.