v1.1.9

1.1.9 (2026-08-09)

Bug Fixes

• cli: don't auto-spawn dashboard for read-only subcommands (#838) (e638ced)
• hardware: match adapter names through OEM-mangled trademark symbols (edf7af2)
• hardware: read 64-bit registry VRAM and full DIMM total for AMD APU UMA carveouts (#817) (c57afd3), closes #810
• hardware: read true Windows VRAM from the display driver registry (#831) (bd7f07a), closes #830
• hardware: skip legacy Intel HD GPUs (#841) (0943a50)
• normalize llama.cpp GGUF model IDs (#833) (03c487d)
• plan: grade run_paths fit_level against the machine's real memory pools (#813) (6e72c22)
• plan: honour CalcConfig so plan and fit agree under tuned settings (#834) (db72ff9)
• plan: route speed estimates through fit.rs so MoE models aren't 4x underestimated (#807) (e7683c1)
• resolve generic AMD/ATI APU identity (#842) (0474cd7)
• share: give the declared memory tier a floor below 8 GB (#818) (34b050a), closes #816
• share: strip absolute GGUF paths from the pending store model field (#820) (777f2c4)

What's Changed

• [Automated] Weekly benchmark cache refresh by @github-actions[bot] in #828
• fix(share): give the declared memory tier a floor below 8 GB by @Akciali in #818
• fix(hardware): read true Windows VRAM from the display driver registry by @AlexsJones in #831
• chore(deps): bump astral-sh/setup-uv from 8.2.0 to 9.0.0 by @dependabot[bot] in #826
• fix: normalize llama.cpp GGUF model IDs by @AlexsJones in #833
• fix(hardware): read 64-bit registry VRAM and full DIMM total for AMD APU UMA carveouts by @AlexsJones in #817
• chore(deps): bump docker/login-action from 4.4.0 to 4.5.1 by @dependabot[bot] in #825
• fix(plan): route speed estimates through fit.rs so MoE models aren't 4x underestimated by @marzmesas in #807
• fix(plan): grade run_paths fit_level against the machine's real memory pools by @marzmesas in #813
• community: add Apple M4 Pro (40 models, llama.cpp Metal) by @Akciali in #812
• bench: community results for apple-m5-pro by @umigliore in #811
• chore(deps): bump clap from 4.6.1 to 4.6.4 by @dependabot[bot] in #800
• fix: resolve generic AMD/ATI Strix Halo identity by @hogeheer499-commits in #842
• fix(hardware): skip legacy Intel HD GPUs on Windows by @SulimanAbdulrazzaq in #841
• fix(plan): honour CalcConfig so plan and fit agree under tuned settings by @marzmesas in #834
• bench: community results for intel-alder-lake-p-gt2-iris-xe-graphics-integrated by @datawookie in #836
• chore(deps): bump http-body-util from 0.1.3 to 0.1.4 by @dependabot[bot] in #802
• chore(deps): bump jsonschema from 0.46.5 to 0.47.0 by @dependabot[bot] in #801
• fix(cli): don't auto-spawn a detached dashboard server for read-only subcommands by @AlexsJones in #838
• chore(deps): bump docker/setup-buildx-action from 4.1.0 to 4.2.0 by @dependabot[bot] in #765
• fix(share): strip absolute GGUF paths from the pending store model field by @Akciali in #820
• bench: community results for nvidia-geforce-rtx-5070 by @DaggyBaggy in #794
• chore(deps): bump sysinfo from 0.39.3 to 0.39.6 by @dependabot[bot] in #764
• bench: community results for nvidia-rtx-5880-ada-generation by @liujiaxiu666 in #796
• chore(deps): bump which from 8.0.4 to 8.0.5 by @dependabot[bot] in #798
• chore(main): release 1.1.9 by @AlexsJones in #832

New Contributors

• @marzmesas made their first contribution in #807
• @umigliore made their first contribution in #811
• @hogeheer499-commits made their first contribution in #842
• @datawookie made their first contribution in #836
• @DaggyBaggy made their first contribution in #794
• @liujiaxiu666 made their first contribution in #796

Full Changelog: v1.1.8...v1.1.9