Pyyan / Compare / Beam 1.0 vs DeepSeek V4.1 Flash
Beam 1.0 vs DeepSeek V4.1 Flash
2 of 5Open-Weight Models · verified 26 Oct 2026
RBeam 1.0Reflectioncurrent
DeepSeek V4.1 FlashDeepSeekcurrent | Specification | Beam 1.0 | DeepSeek V4.1 Flash |
|---|
| Summary | 501B under Apache 2.0, with nothing asked in return. | Twice the parameters of the Flash it replaces, and cheaper to run. |
|---|
| Context | 1M | 1M |
|---|
| Input ($/Mtok) | $0.25 | $0.3 |
|---|
| Output ($/Mtok) | $1 | $1.2 |
|---|
| Licence | Apache 2.0 | MIT |
|---|
| Max output | 131K | 256K |
|---|
| Parameters | 501B total, 23B active | 552B total, 8B / 16B active |
|---|
| Released | 16 Oct 2026 | 10 Sep 2026, refreshed 17 Oct |
|---|
| Model ID | reflection/beam-1.0-501b-a23b | deepseek-flash |
|---|
| Category | Open-Weight Models | Open-Weight Models |
|---|
| Official | Reflection ↗ | DeepSeek ↗ |
|---|
Highlighted rows are where these differ.
Beam 1.0
- Apache 2.0, with no attribution clause and no revenue trigger, unlike Meta's Muse licence
- 52 layers, 256 experts, a 1M token context and a knowledge cutoff of 30 June 2026
- Reflection claims comparable answers to leading open models with roughly three times fewer reasoning tokens
- Artificial Analysis scores it 30 on the intelligence index, 19th of 120 models
Best for open weights with no licence conditions attached.
Full spec sheet →
DeepSeek V4.1 Flash
- 552B total, 8B active while reading the prompt and 16B while generating, with a 1M token context
- DeepSeek says it beats V4-Pro; V4-Pro requests route to it from 14 September 2026
- MIT licence, native image input, and a KV cache about a quarter the size of V4-Flash's
- Refreshed on 17 October as V4.1-Flash-1017, still MIT, improving on all eleven benchmarks DeepSeek published
Best for cheap agentic coding with images.
Full spec sheet →