melvincarvalho on Nostr: Qwen3-30B-A3B runs on NVIDIA 4090 at 100 tokens/s! It's a very good model, which will ...
Qwen3-30B-A3B runs on NVIDIA 4090 at 100 tokens/s!
It's a very good model, which will only get better.
While everyone is waiting on deepseek r2 this little beauty might be all you need for consumer grade AI.
Published at
2025-04-30 18:01:35 GMTEvent JSON
{
"id": "f7c21e0e83e7d6af280efdeb0430787bd5d5ed55aeb9e2619a06f8b871a1d715",
"pubkey": "de7ecd1e2976a6adb2ffa5f4db81a7d812c8bb6698aa00dcf1e76adb55efd645",
"created_at": 1746036095,
"kind": 1,
"tags": [
[
"client",
"Ditto",
"31990:15b68d319a088a9b0c6853d2232aff0d69c8c58f0dccceabfb9a82bd4fd19c58:ditto",
"wss://ditto.pub/relay"
]
],
"content": "Qwen3-30B-A3B runs on NVIDIA 4090 at 100 tokens/s!\n\nIt's a very good model, which will only get better.\n\nWhile everyone is waiting on deepseek r2 this little beauty might be all you need for consumer grade AI.",
"sig": "0dcbbfc27ed027ba78cafbfcb71951cf662048bc332d372b193ea2c3edf3d51122f110f77b37e0f8c5148153546015b9ea2d15f4897c14d144b3c8c37355b457"
}
