{"schemaVersion":1,"id":"mlx-lm","revision":"2026-09-15.1","name":"MLX-LM","monogram":"M","category":"Inference library","openness":"Open source","license":"MIT runtime source; model weights have separate licenses","color":"#75c7d8","summary":"Python generation tools for MLX-compatible models on Apple Silicon.","website":"https://github.com/ml-explore/mlx-lm","repository":"https://github.com/ml-explore/mlx-lm","environment":"Local Apple Silicon execution with MLX; model conversion and server configuration remain operator-owned.","contributor":"Testedknowhow model runtime research","affiliation":"Commissioned by a Nimbalyst maintainer. Documentation synthesis; no independent review or product execution.","contentLicense":"Unlicensed prototype; redistribution decision pending","reviewStatus":"unreviewed","lastChecked":"2026-09-15","expiresAt":"2026-10-15","applicability":{"versions":["Repository revision 872ae88d1fac77350db23c8c04fe8dd372a9e3e8; living documentation separately dated."],"environment":"Local Apple Silicon execution with MLX; model conversion and server configuration remain operator-owned.","excludes":["Runtime verification and performance rankings","Model weights license or quality evaluation","Cloud inference, remote hosts and external tools","Unlisted model architectures, builds and hardware"]},"claims":[{"id":"mlx-lm-rt-local","capability":"rt-local","statement":"Provides local text generation with MLX.","method":"documented","outcome":"claimed","scope":"Model files must be acquired before an offline run.","evidenceIds":["mlx-lm-readme"],"contradictions":[],"support":"yes"},{"id":"mlx-lm-rt-apple-silicon","capability":"rt-apple-silicon","statement":"Targets Apple Silicon through the MLX framework.","method":"documented","outcome":"claimed","scope":"Supported Apple hardware and MLX installation; speed unmeasured.","evidenceIds":["mlx-lm-readme"],"contradictions":[],"support":"yes"},{"id":"mlx-lm-rt-mlx","capability":"rt-mlx","statement":"Loads MLX-compatible model repositories or local conversions.","method":"documented","outcome":"claimed","scope":"Supported architectures and MLX-compatible weight layouts.","evidenceIds":["mlx-lm-readme"],"contradictions":[],"support":"yes"},{"id":"mlx-lm-rt-quantized","capability":"rt-quantized","statement":"Provides conversion to quantized model weights.","method":"documented","outcome":"claimed","scope":"README conversion example uses 4-bit weights; not equivalent to GGUF 4-bit artifacts.","evidenceIds":["mlx-lm-readme"],"contradictions":[],"support":"yes"},{"id":"mlx-lm-rt-streaming","capability":"rt-streaming","statement":"Provides incremental output through stream_generate.","method":"documented","outcome":"claimed","scope":"Python streaming interface; no blanket HTTP API parity claim.","evidenceIds":["mlx-lm-readme"],"contradictions":[],"support":"yes"},{"id":"mlx-lm-rt-prompt-cache","capability":"rt-prompt-cache","statement":"Provides prompt caching for repeated long prompts.","method":"documented","outcome":"claimed","scope":"mlx_lm.cache_prompt and generation cache options; capture cache state during benchmarks.","evidenceIds":["mlx-lm-readme"],"contradictions":[],"support":"yes"},{"id":"mlx-lm-rt-http","capability":"rt-http","statement":"Declares a configurable HTTP server interface.","method":"documented","outcome":"claimed","scope":"Repository-declared host and port interface; not executed.","evidenceIds":["mlx-lm-server"],"contradictions":[],"support":"yes"},{"id":"mlx-lm-rt-openai","capability":"rt-openai","statement":"Declares a /v1/chat/completions route.","method":"documented","outcome":"claimed","scope":"Repository-declared route only; protocol compatibility and client behavior untested.","evidenceIds":["mlx-lm-server"],"contradictions":[],"support":"yes"}],"evidence":[{"id":"mlx-lm-readme","url":"https://github.com/ml-explore/mlx-lm/blob/872ae88d1fac77350db23c8c04fe8dd372a9e3e8/README.md","title":"MLX-LM documentation","locator":"Quick Start; Python API; quantization; long prompts and generations","capturedAt":"2026-09-15","kind":"repository","revision":"872ae88d1fac77350db23c8c04fe8dd372a9e3e8","redistribution":"Link and original paraphrase only; upstream rights retained."},{"id":"mlx-lm-server","url":"https://github.com/ml-explore/mlx-lm/blob/872ae88d1fac77350db23c8c04fe8dd372a9e3e8/mlx_lm/server.py","title":"MLX-LM server interface","locator":"main argument parser: MLX Http Server, --host, --port; /v1/chat/completions route","capturedAt":"2026-09-15","kind":"repository","revision":"872ae88d1fac77350db23c8c04fe8dd372a9e3e8","redistribution":"Link and original paraphrase only; upstream rights retained."},{"id":"mlx-lm-license","url":"https://github.com/ml-explore/mlx-lm/blob/872ae88d1fac77350db23c8c04fe8dd372a9e3e8/LICENSE","title":"MLX-LM license","locator":"MIT License","capturedAt":"2026-09-15","kind":"repository","revision":"872ae88d1fac77350db23c8c04fe8dd372a9e3e8","redistribution":"Link and original paraphrase only; upstream rights retained."}],"gaps":["Which pinned model, quantization, context length and runtime versions fit the selected Mac?","What are measured latency, throughput, peak memory and swap under one and four concurrent requests?","Concurrency and JSON schema output are Unknown in this research package; no cited absence is asserted."],"invalidationTriggers":["Runtime, bundled engine, backend, model format or license changes.","Documentation changes affecting the cited configuration."],"verificationPlan":["Choose hardware and obtain explicit approval for installations, downloads and execution resources.","Pin runtime and engine versions, model artifact hashes, tokenizer, template, quantization and inference settings.","Run the local-model-mac investigation with synthetic prompts and capture raw measurements; do not record documentation as execution."],"attestations":[],"space":"model-runtimes"}