Evo 2 7B
- Family
- StripedHyena 2
- Scale
- 7B dense
- Context
- Up to 1M bases
gpt-oss 20B
- Depth
- 24 layers
- Scale
- 20.91B / 3.61B active
- Context
- 131,072 tokens
DeepSeek V3
- Depth
- 61 layers
- Scale
- 671B / 37B active
- Cache idea
- 512 content + 64 position
Qwen3-Next 80B-A3B
- Depth
- 48 layers
- Scale
- 80B / ~3B active
- Context
- 262,144 native