From da132cd12831fbb21365c6db29850d2165c6c836 Mon Sep 17 00:00:00 2001 From: Andrew Anderson Date: Thu, 5 Nov 2026 10:26:54 -0500 Subject: [PATCH] =?UTF-8?q?=E2=9C=A8=20homepage:=20add=20Apple=20MLX=20to?= =?UTF-8?q?=20the=20integrations=20strip?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Adds Apple MLX to the Inference engines list on the homepage, using the MLX project's own MIT-licensed wordmark (ml-explore/mlx docs/src/_static/mlx_logo.png, cropped and downscaled to 64x64 like the other PNG marks). Mirrors the name into llms.txt and the JSON-LD keywords/featureList so the parity tests stay green. Hive v5 supports Apple MLX via the Pi backend with an mlx/ provider (src/docs/integrations.md, src/docs/contributor-relay.md, bin/mlx-backend.js). Signed-off-by: Andrew Anderson --- assets/integrations/mlx.png | Bin 0 -> 1645 bytes index.html | 3 ++- llms.txt | 2 +- 3 files changed, 3 insertions(+), 2 deletions(-) create mode 100644 assets/integrations/mlx.png diff --git a/assets/integrations/mlx.png b/assets/integrations/mlx.png new file mode 100644 index 0000000000000000000000000000000000000000..59b747915b6ad2845db0a03ae142cbf6b57e29d3 GIT binary patch literal 1645 zcmV-z29o)SP)000IwNklfRSZezMh+G;O?iZ}Hj3a=LxLHbe>i5WINu;6Z2$tqAEs z6t5zJ2YcvEh_9%iO|ngzCQTPN*`1m9@U-KaCb2IKh`c+8u$i6NozL@ppJ$#OETu$$ zawF&uZvj965C8-K0YCr{00aO5KmZT`1ONfR?`lk4?ff{#7tc*C#zKgP4F);zNI(l+tWu?_>!FxI9!^6X|SWHUkcOB>a>({Sf zYbhmTY;<b{Gpj^n(2`_}*HM~@zzI(4d>R31Edu(Y(KI+Mwy8fn|M|5m9~>L@wr z#2Clp@h@M#0G<${R4PRx5m=lWQ>j$y^$o@-z>JNJ5klY=QOgR4!<16S7-Ni5`s~@W zTCK(y^S>M1_5S_)ckkZ8Fiq1arO{~gM_M3+Qc5Y+bsYwDl`qCp%6L4UNF=0`gb>D< z<2co7RWW0XTdmgG+M1>@aJXR@hG8fd5&z;i=bUrT^AJLnO6B?U=RoWqgAkgZpNBDg zyFtI-)ge=SALV&OtW2scipBqSz=g!Q`fJK!zDW%P3 z^U0GZ3TE#OG;NwDSPEn8y6)=gDqP#PU9Z;}-4c&!nwqBhjnWx_bB-~7^XAQm4<862 zuvYke_3G8<&z}jU{Q>|>8-@{!#ejiQy1u^dI1ZR!Eh!$4`^D200OEsE>bma2!h+w} z7~?IsV2m+V&e=NyLMWY1tI!lemdoW@t!CRcNS{)gPN#QnEM%ucj~qF2>Cz=|!u-O# z>$;Ru$fuTNy?pr+W9)g})vH&N$z&ger0aSzncQ{)X*3#zLV^0k6AFcPMpOqAYMOTI z)-8bk_|M1JuU{jCT-QYiEiNuL8VyKU*=)AiZ1!zK#TX%U;J|@_fdR0c>gSwCqtT(E zp*Hb!1=y(9FI>1VK0Xe{n3g0iVVXDhr^&KN@=N7 z0&DIH0G={AIcW?IN-1sIzIX2)B-7)^kDobnX7`%_*s^15Gk>X65<;qyA(P3Nrm65; z*X>vrb}c1@5Q#)$;_ttK2g+T9(Dd{);O)`_x~>yKLZJ}Z(fsx1`ue&GDj`HFmBJWD zqtX5Q_e0&q7^~H4ki`8AtaoD0X0x-ivtTXGc{m(~T7NeUX5GDe_wnP$LWm#b7a>B3 zv9Yn)*;zseY>Spe{j_&-0*UAcUxT zQ%b36+VJpjp-^Cq5ki{H=IZKdDwS$4r=^sR(*$XL|Nb31VWHQrUl&3+jw6J)dGjV{ zdV7kqHN|muik|;0^tKiXh2WymXjH3JSgz-Ji9{kC4y%s=Ddo`6P$UxZJP+{7<+5$t z%AZV?4~GvQhI(OSWCU)5BClP$Ha9odY&K2Pym;{<#u&`saSE(#e0*HJzGIpeLL54D z2nJNERfrQQ<-ou|GMQ`(Fc8wWZ{NVNLWs4swSD{c!5FMQG$B?LPg_=n%+Pl9&~Zvz ze~%agL-@m^cVh6@7b*;XeJbHk`F&6S{XACbahRiSb?fhOF7(R^^#6V?6r7C)00BS% r5C8-K0YCr{00aO5KmZT`{I2#d$e80GURwRG00000NkvXXu0mjfHtroH literal 0 HcmV?d00001 diff --git a/index.html b/index.html index f45a94c..9e20e20 100644 --- a/index.html +++ b/index.html @@ -26,7 +26,7 @@ - + @@ -658,6 +658,7 @@

Inference engines

  • llm-dllm-d project · backend: llm-d
  • LiteLLMBerriAI · backend: litellm
  • IBM watsonx.aiIBM · backend: watsonx
  • +
  • Apple MLXMLX-LM on Apple Silicon · via Pi: mlx/<model>
  • diff --git a/llms.txt b/llms.txt index 45f816f..bb6b4ce 100644 --- a/llms.txt +++ b/llms.txt @@ -21,7 +21,7 @@ Hive Commons also maintains Spektacular (Spek), a spec-driven workflow tool; the Hive supports agent CLIs including Claude Code, GitHub Copilot CLI, Goose, OpenAI Codex CLI, Pi, IBM Bob, Aider, Gemini CLI, Google Antigravity CLI, OpenCode, Kilo Code, Muse Code and Oh My Pi. -Hive supports inference engines and gateways including vLLM, llm-d, LiteLLM, IBM watsonx.ai, OpenRouter, Anthropic, OpenAI and DeepSeek. +Hive supports inference engines and gateways including vLLM, llm-d, LiteLLM, IBM watsonx.ai, Apple MLX, OpenRouter, Anthropic, OpenAI and DeepSeek. See all integrations: https://docs.hivecommons.dev/docs/hive/integrations