6 alternatives to airllm

lyogavin/airllm — AirLLM 2.8T LLM inference with single 4GB GPU (41,451 stars, Apache-2.0, ai).

These are the closest projects in the JustClaw-It index that we serve licence-cleared code from, ranked by relevance and by our code-quality score (1–10, the mean over each repo's graded components: completeness, code quality, deployability, reusability, strategic value). JustClaw-It does not serve code from lyogavin/airllm itself (licence class: permissive), so these alternatives are the licence-cleared route.

The alternatives in detail

• vllm-project/vllm
  A high-throughput and memory-efficient inference and serving engine for LLMs
  score 6.9/10 · Apache-2.0 permissive · 94,666 stars · 3,001 components · search them · alternatives
• bentoml/bentoml
  The easiest way to serve AI apps and models - Build Model Inference APIs, Job queues, LLM apps, Multi-model pipelines, and more!
  score 7.7/10 · Apache-2.0 permissive · 8,917 stars · 230 components · search them
• bytedance/infinistore
  KV cache store for distributed LLM inference
  score 5.7/10 · Apache-2.0 permissive · 449 stars · 23 components · search them
• nousresearch/nous-llama.cpp
  LLM inference in C/C++ - nous research
  score 4.4/10 · MIT permissive · 8 stars · 71 components · search them
• ggml-org/llama.cpp
  LLM inference in C/C++
  score 6.7/10 · MIT permissive · 133,192 stars · 1,798 components · search them
• nousresearch/speculators
  A unified library for building, evaluating, and storing speculative decoding algorithms for LLM inference in vLLM
  score 4.8/10 · Apache-2.0 permissive · 3 stars · 77 components · search them

Licence data: JustClaw-It's per-repo licence policy. Is airllm still maintained? Check it. Licence details: licence checker. Ranking data from 2026-11-04T22:41:13Z.