ricursiveAI models
3 ricursiveAI models from ricursiveAI on Hugging Face, from 4.0B to 64.8B parameters, published by ricursiveAI in BF16, with contexts from 8K to 2M tokens. The smallest official model, Ripple-65B-Preview-DFlash2, needs about 9.0 GB of VRAM at its published precision; the cheapest live fit is RTX 4070 at $0.154/hr.
Pick a size
One row per official ricursiveAI size: the VRAM it needs at each precision, the cheapest GPU that holds it today, and the KV cache for a 32K-token context.
| Model | Parameters | Native VRAM | FP8 VRAM | INT4 VRAM | Live GPU fit (native) | Est. $/hr | KV cache at 32K |
|---|---|---|---|---|---|---|---|
| Ripple-65B-Preview-DFlash2 | 4.0B | 9.0 GB | 4.5 GB | 2.3 GB | RTX 4070 | $0.154/hr | not published for this architecture |
| Ripple-65B-Preview | 64.8B | 145 GB | 72.4 GB | 36.2 GB | RTX 3090 × 7 | $1.13/hr | 2.3 GB |
VRAM is weights times a flat 1.2 overhead; the KV cache is a separate per-model figure at 16-bit, shown where the architecture is published. See the methodology.
Official models (3)
| Model | Parameters | Published as | VRAM needed | Live GPU fit | Est. $/hr |
|---|---|---|---|---|---|
| Ripple-65B-Preview-DFlash2 | 4.0B | BF16 | 9.0 GB | RTX 4070 | $0.154/hr |
| Ripple-65B-Base-Preview | 64.8B | BF16 | 145 GB | RTX 3090 × 7 | $1.13/hr |
| Ripple-65B-Preview | 64.8B | BF16 | 145 GB | RTX 3090 × 7 | $1.13/hr |
VRAM is for the precision the model is published in: the weights times a flat 1.2 overhead, with the KV cache not included. See the methodology. The fit shown is the lowest-priced single GPU type that holds the model at that precision, or the lowest-priced multi-GPU set (up to 8) when none does.
