vMLX 1.5.65
vMLX 1.5.65
Highlights:
- Moves the MiniMax-M3 compatibility release 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.
- MiniMax-M3 UI/API coverage includes 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.
- Cache/memory UI now shows effective paged-cache capacity and marks memory fields ignored when paged cache is active, so CLI and GUI users see the real cache-pressure knobs.
- Gemma 4 VL scoped rows remain in the release gate; Gemma audio stays out of scope for this release.
Verification:
- vMLX main source commit: 43a5f460d.
- v1.5.65 source tag: 43a5f460d.
- Scoped 1.5.65 MM3/Gemma preflight before packaging: status=pass.
- Bundled Python verification passed for Sequoia and Tahoe: vmlx_engine 1.5.65, jang-tools, MLX, mlx-lm, mlx-vlm, MiniMax/Gemma-related imports, and bundled source parity checks.
- Sequoia and Tahoe DMGs are Developer ID signed by ShieldStack LLC (55KGF2S5AY), Apple notarized, stapled, Gatekeeper accepted, and verified by the release DMG verifier.
Notarization:
- Sequoia notary id: 9d76633e-abd1-4e60-81c5-2b526d3da4dc.
- Tahoe notary id: 5b6c3d94-c840-4beb-af42-26aa155a07e9.
Artifacts:
- Sequoia DMG SHA256: 7c9984b9fdafeffbe470b2d7396e43bfd54138b6cc832e923a1445a245bd5ed3
- Tahoe DMG SHA256: a9bc04daa894da6552906d09f9ec4a7056eea41041f39178ccb6981ffcad21b4
Scope:
- This is a scoped MM3 + Gemma 4 VL compatibility release. Broad non-scoped model-family rows and extra diagnostic cache-mode hardening remain next-release work.