Reflection vs 
Sapiens AI
Verdict
- Sapiens AI's flagship leads: Agnes 2.5 Pro Beta at 59.1 vs Beam 1.0 at 56.4.
- Sapiens AI has the better model in 4 of 4 price bands.
- Reflection's fastest ranked model (Beam 1.0, 303 tok/s) beats Sapiens AI's (Agnes 2.5 Pro Beta, 111 tok/s).
- Reflection publishes more open-weights models (2 vs 0).
Best model per price band
Each provider’s highest-index model below the price ceiling. Purple marks the winner of the band.
| Price band | Reflection | Sapiens AI |
|---|---|---|
| Under $0.50/M | – | Agnes 2.5 Pro Beta 59.1$0.150/M |
| Under $2/M | Beam 1.0 56.4$1.05/M | Agnes 2.5 Pro Beta 59.1$0.150/M |
| Under $10/M | Beam 1.0 56.4$1.05/M | Agnes 2.5 Pro Beta 59.1$0.150/M |
| Any price | Beam 1.0 56.4$1.05/M | Agnes 2.5 Pro Beta 59.1$0.150/M |
Line-up
| Metric | Reflection | Sapiens AI |
|---|---|---|
| Flagship index | 56.4 | 59.1 |
| Ranked models | 1 | 1 |
| Open-weights models | 2 | 0 |
| Cheapest ranked, $/M | $1.05 | $0.150 |
| Fastest ranked, tok/s | 303 | 111 |
| Release cadence, days | – | – |
| Last release | – | – |
| Next expected | – | – |
Data as of 18 Oct 2026. See Reflection and Sapiens AI in full.