Skip to content

vMLX 1.5.65

Choose a tag to compare

View all tags
@jjang-ai jjang-ai released this 18 Jun 20:13
· 2051 commits to main since this release
v1.5.65
d47ff6b

vMLX 1.5.65

Source release for the scoped MM3 + Gemma 4 VL compatibility update.

Highlights:

  • MiniMax-M3 compatibility release moved to a new updater-visible version after 1.5.64 shipped.
  • MiniMax-M3 startup keeps prefix cache enabled, paged cache disabled, disk prompt cache enabled, JIT disabled, native MSA/Lightning sparse cache storage first-class, and generic TurboQuant KV disabled.
  • UI/API paths cover MM3 reasoning off/on/auto, visible content emission, exact tool calls, Responses, Chat Completions, Anthropic-compatible, Ollama-compatible, streaming, VL image input, and prefix-cache reuse telemetry.
  • latest.json now points to the notarized 1.5.65 Sequoia/Tahoe DMGs.

Verification:

  • vMLX main commit: 43a5f46.
  • v1.5.65 tag source commit: 43a5f46.
  • Scoped 1.5.65 MM3/Gemma preflight before packaging: status=pass.
  • Sequoia/Tahoe DMGs are published from the mlxstudio release, Developer ID signed, Apple notarized, stapled, Gatekeeper accepted, and verified by the release DMG verifier.

Download DMGs:

DMG SHA256:

  • Sequoia: 7c9984b9fdafeffbe470b2d7396e43bfd54138b6cc832e923a1445a245bd5ed3
  • Tahoe: a9bc04daa894da6552906d09f9ec4a7056eea41041f39178ccb6981ffcad21b4
Assets 2
Loading