Solar
29 opinions tracked this week · 7-day rolling window
30-day experience score
By experience dimension
| General text | 40 | n=7 | |
|---|---|---|---|
| Coding | 32 | n=9 | |
| Reasoning | 47 | n=7 | |
| Image / vision | not enough data | n=0 | |
| Roleplay / creative | not enough data | n=2 | |
| Safety & refusals | not enough data | n=0 | |
| Speed & latency | not enough data | n=4 | |
| Local deploy | not enough data | n=0 |
Top positive drivers
- strong reasoning Reasoning ×2
- speed quality Coding ×1
- performance issue Reasoning ×1
- creativity up Roleplay / creative ×1
- efficient consumption Speed & latency ×1
Top negative drivers
- performance issue Coding ×2
- code slow sqlite locking broken Coding ×1
- hallucination invention Coding ×1
- overengineering simple task Coding ×1
- overengineering tendency Coding ×1
What users are saying
Selected from the latest 7 days, prioritizing recency, evidence quality, confidence, and engagement.
🔴 “SOL xhigh cannot stop itself from finding edge case of edge cases of edge cases and get lost in the rabbit hole. It is pretty frustrating. You can sense the potential of the model but to get it to stay on track has been a bit of a coin toss”
via reddit · reasoning · en · 2026-07-15
🟢 “try Sol medium... its not going to be trained to constantly loop analysing everything for millions of tokens on end”
via reddit · reasoning · en · 2026-07-14
🔴 “I used sol medium to add a workout tracking feature to my personal os, left it running on /goal came back to 30k LOC”
via reddit · coding · en · 2026-07-14
🔴 “Never ever never go above high for execution, medium for a well written plan is way more then enough. It's not about token economy, it's actually better software.”
via reddit · coding · en · 2026-07-14
🔴 “inventing hypothetical overrides to guard against is a fascinating failure mode, like the tool started debugging its own imagination”
via reddit · coding · en · 2026-07-14