Model reference · open weights
Dream is an open-weight language model from Dream-org. Dream-v0-Instruct-7B (BF16) weighs 15.2 GB; the smallest configuration that runs it is 2× RTX 3060 12 GB.
For self-hosting, compare this build's requirements below with the GPU you plan to rent.
What it is
| Released by | Dream-org |
|---|---|
| Released | 2025-04-03 |
| Parameters | 7.6B |
| VRAM | 15.2 GB for the weights |
What it runs on
| Card | Requests at once | Context max | Memory | |
|---|---|---|---|---|
| 8K each | 32K each | |||
| RTX 3060 12 GB … RTX 4060 Ti 16 GB 2 smaller cards | ||||
| RTX 3090 24 GB | 15 | 3 | 125K | 23.4 GB |
| RTX 4090 24 GB | 15 | 3 | 124K | 23.4 GB |
| RTX 5090 32 GB | 31 | 7 | all 128K | 31.0 GB |
| L40S 48 GB | 59 | 14 | all 128K | 44.0 GB |
| A100 80 GB | 132 | 33 | all 128K | 78.2 GB |
| H100 80 GB | 122 | 30 | all 128K | 78.1 GB |
| RTX PRO 6000 Blackwell 96 GB | 156 | 39 | all 128K | 93.8 GB |
| DGX Spark (GB10) 128 GB unified | 184 | 46 | all 128K | 107 GB |
| H200 141 GB | 250 | 62 | all 128K | 138 GB |
| B200 180 GB | 329 | 82 | all 128K | 176 GB |
| 2× RTX 3060 12 GB tensor parallel | 13 | 3 | 109K | 11.6 GB a card |
| 2× RTX 4060 Ti 16 GB tensor parallel | 29 | 7 | all 128K | 15.4 GB a card |
| Requests at once | 8K tokens each | 32K tokens each |
|---|---|---|
| 1 | 16.5 GB | 17.9 GB |
| 5 | 18.4 GB | 25.4 GB |
| 8 | 19.8 GB | 31.1 GB |
| 16 | 23.5 GB | 46.1 GB |
| 32 | 31.1 GB | 76.2 GB |
| 64 | 46.1 GB | 136 GB |
One card, with vLLM's small-card settings.
From the model card
This is the instruct model of Dream 7B, which is an open diffusion large language model with top-tier performance. More details about the model and usage can be found in the blog and github bellow:
Quoted from the model card on Hugging Face. The full card is behind the Hugging Face link above.