NVIDIA vs
Reflection
Verdict
- NVIDIA's flagship leads: Nemotron 3 Ultra 550B A55B at 55.4 vs Beam 1.0 501B A23B at 54.2.
- NVIDIA has the better model in 3 of 4 price bands.
- Reflection's fastest ranked model (Beam 1.0, 288 tok/s) beats NVIDIA's (Nemotron 3.5 Super VL, 271 tok/s).
- NVIDIA publishes more open-weights models (21 vs 2).
Best model per price band
Each provider’s highest-index model below the price ceiling. Purple marks the winner of the band.
| Price band | NVIDIA | Reflection |
|---|---|---|
| Under $0.50/M | Nemotron 3 Super 120B A12B thinking 51.1$0.450/M | Beam 1.0 501B A23B 54.2$0.475/M |
| Under $2/M | Nemotron 3 Ultra 550B A55B 55.4$1.20/M | Beam 1.0 501B A23B 54.2$0.475/M |
| Under $10/M | Nemotron 3 Ultra 550B A55B 55.4$1.20/M | Beam 1.0 501B A23B 54.2$0.475/M |
| Any price | Nemotron 3 Ultra 550B A55B 55.4$1.20/M | Beam 1.0 501B A23B 54.2$0.475/M |
Line-up
| Metric | NVIDIA | Reflection |
|---|---|---|
| Flagship index | 55.4 | 54.2 |
| Ranked models | 11 | 2 |
| Open-weights models | 21 | 2 |
| Cheapest ranked, $/M | $0.088 | $0.475 |
| Fastest ranked, tok/s | 271 | 288 |
| Release cadence, days | 36 | – |
| Last release | 23 Oct 2026 | – |
| Next expected | 13 Nov 2026 – 12 Dec 2026 | – |
Data as of 16 Nov 2026. See NVIDIA and Reflection in full.