ReVolta
Inference server that runs Qwen3.8 Flash-Next on multi-GPU V100 servers via OpenAI-compatible API.
Not verified yetKnown vulnerabilities
How far pickfit got
pickfit hasn’t verified ReVolta yet, so there are no tested install steps.
What follows comes from its README only.
Use it
Before you install: known vulnerabilities
pickfit found 10 known high or critical vulnerabilities in the packages a fresh install brings (OSV.dev, 3 Nov 2026). Weigh this before you install it.
crates:openssl 0.10.76rust-openssl has incorrect bounds assertion in aes key wrap (GHSA-8c75-8mhr-p7r9, high, fixed in 0.10.78)crates:pulldown-cmark 0.13.3Exponential event growth with `ENABLE_WIKILINKS` (RUSTSEC-2026-0414, high, fixed in 0.13.5)
See 8 more
crates:openssl 0.10.76— GHSA-ghm9-cr32-g9qj (high), fixed in 0.10.78crates:openssl 0.10.76— GHSA-hppc-g8h3-xhp3 (high), fixed in 0.10.78crates:openssl 0.10.76— GHSA-pqf5-4pqq-29f5 (high), fixed in 0.10.78crates:openssl 0.10.76— GHSA-xp3w-r5p5-63rr (high), fixed in 0.10.79crates:pulldown-cmark 0.13.3— RUSTSEC-2026-0433 (high), fixed in 0.13.6crates:quinn-proto 0.11.14— GHSA-4w2j-m93h-cj5j (high), fixed in 0.11.15crates:quinn-proto 0.11.14— RUSTSEC-2026-0340 (high), fixed in 0.11.18crates:rustls-webpki 0.103.9— GHSA-82j2-j2ch-gfr8 (high), fixed in 0.103.13
You need first
- Runtimes:
Python 3.12 - A GPU is required
- At least 32 GB of RAM
What it does
pickfit hasn’t run it yet, so these come from the README only.
- Run inference on text, images and videoNot tested
- Serve the model through an OpenAI-compatible APINot tested
- Offload the KV cache to CPU memory or diskNot tested
- Use speculative decoding with MTP or DFlash2Not tested
Good for
- serve Qwen3.8 Flash-Next for text, images and video via OpenAI-compatible API
- recover cached long-context prefixes from CPU memory or disk
- serve 512K-token prompts on V100 GPUs