Alternatives
Products that do what Bhumi–OSS Python Library w Rust Underhead for 2.5x Faster LLM Inference does
Read the full blogpost at https://rach.codes/blog/Introducing-Bhumi (click on reader to see the technical breakdown!) AI inference should be fast, but in practice it’s painfully slow. Inference bottlenecks slow down LLM-powered chatbots and AI workflows everywhere. I built Bhumi to fix that. Bhumi is a Python library designed for developers, yet its performance-critical core is implemented in Rust (via PyO3) for near-native speed. This hybrid approach delivers up to 2.5x faster response times across providers like OpenAI, Anthropic, and Gemini—without changing the…
- 1WM
We wrote our inference engine on Rust, it is faster than llama cpp in all of the use cases. Your feedback is very welcomed. Written from scratch with idea that you can add support of any kernel and platform.
2025 · github.com
- 2OS
Hi HN, I built a specialized inference engine for running 4-bit Gemma 4 26B-A4B-IT on any M-series Mac using about 2 GB of RAM. It is called TurboFieldfare and is written in Swift and Metal. I have always adored on-device AI. It feels like magic that you can run a powerful NN on your Mac or iPhone. So I wanted to push the limits a bit and run a model whose weights don’t fit in memory. The model’s 4-bit quantized weights occupy roughly 14 GB, which makes running it with conventional inference tools almost impossible on an 8 GB or even 16 GB Mac once the OS, applications, and KV cache are…
Jul 2026 · github.com
- 3SU
Here's a project I've been working on for the last few months. It's a new (I think) algorithm, that allows to adjust smoothly - and in real time - how many calculations you'd like to do during inference of an LLM model. It seems that it's possible to do just 20-25% of weight multiplications instead of all of them, and still get good inference results. I implemented it to run on M1/M2/M3 GPU. The mmul approximation itself can be pushed to run 2x fast before the quality of output collapses. The inference speed is just a bit faster than Llama.cpp's, because the rest of implementation…
2024 · asciinema.org
- 4FF
A few months ago, I benchmarked FastAPI on an i9 MacBook Pro. I couldn't believe my eyes. A primary REST endpoint to `sum` two integers took 6 milliseconds to evaluate. It is okay if you are targeting a server in another city, but it should be less when your client and server apps are running on the same machine. FastAPI would have bottleneck-ed the inference of our lightweight UForm neural networks recently trending on HN under the title "Beating OpenAI CLIP with 100x less data and compute". (Thank you all for the kind words!) So I wrote another library. It has been a while since I have…
2023 · github.com
- 5L3
I spent a lot of time and money on this rather big side project of mine that attempts to replicate the mechanistic interpretability research on proprietary LLMs that was quite popular this year and produced great research papers by Anthropic [1], OpenAI [2] and Deepmind [3]. I am quite proud of this project and since I consider myself the target audience for HackerNews did I think that maybe some of you would appreciate this open research replication as well. Happy to answer any questions or face any feedback. Cheers [1]…
2024 · github.com
- 6
General Compute▲315AI models that run on an inference cloud optimized for speed
May 2026 · generalcompute.com
- 78F
Hi HN! I'm just sharing a project I've been working on during the LLM Efficiency Challenge - you can now finetune Llama with QLoRA 5x faster than Huggingface's original implementation on your own local GPU. Some highlights: 1. Manual autograd engine - hand derived backprop steps. 2. QLoRA / LoRA 80% faster, 50% less memory. 3. All kernels written in OpenAI's Triton language. 4. 0% loss in accuracy - no approximation methods - all exact. 5. No change of hardware necessary. Supports NVIDIA GPUs since 2018+. CUDA 7.5+. 6. Flash Attention support via Xformers. 7. Supports 4bit and 16bit…
2023 · github.com
- 8
- 9
- 10
- 11LA
G'day, HN! I'm one of the maintainers of `llm`. I've been working alongside a trusty group of contributors to bring this project to life, and we're now at a point where we're ready to share it with the world. Large language models (LLMs) are taking the computing world by storm due to their emergent abilities that allow them to perform a wide variety of tasks, including translation, summarization, code generation, and even some degree of reasoning. However, the ecosystem around LLMs is still in its infancy, and it can be difficult to get started with these models. `llm` is a one-stop shop for…
2023 · github.com
- 12AT
I recently built a small open-source tool to benchmark different LLM API endpoints — including OpenAI, Claude, and self-hosted models (like llama.cpp). It runs a configurable number of test requests and reports two key metrics: • First-token latency (ms): How long it takes for the first token to appear • Output speed (tokens/sec): Overall output fluency Demo: https://llmapitest.com/ Code: https://github.com/qjr87/llm-api-test The goal is to provide a simple, visual, and reproducible way to evaluate performance across different LLM providers, including…
2025 · llmapitest.com
- 13MA
Hey HN. I'm Fabien, principal engineer, 25 years shipping production systems (Ruby, Swift, now Rust). I built Moltis because I wanted an AI assistant I could run myself, trust end to end, and make extensible in the Rust way using traits and the type system. It shares some ideas with OpenClaw (same memory approach, Pi-inspired self-extension) but is Rust-native from the ground up. The agent can create its own skills at runtime. Moltis is one Rust binary, 150k lines, ~60MB, web UI included. No Node, no Python, no runtime deps. Multi-provider LLM routing (OpenAI, local GGUF/MLX, Hugging…
Feb 2026 · moltis.org
- 14AE
2025 · github.com
- 15SO
We built SwiftAI, an open-source Swift library that lets you use Apple’s on-device LLMs when available (Apple opened access in June), and fall back to a cloud model when they aren’t available — all without duplicating code. SwiftAI gives you: - A single, model-agnostic API - An agent/tool loop - Strongly-typed structured outputs - Optional chat state Backstory: We started experimenting with Apple’s local models because they’re free (no API calls), private, and work offline. The problem: not all devices support them (older iPhones, Apple Intelligence disabled, low battery, etc.). That…
2025 · github.com
- 16LT
Current AI-assisted CLI tools are often part of larger systems and work better on Linux. I built llm-term to address these. It's a Rust-based tool that compiles into a single binary file. You only need to download the binary, add it to your PATH, and configure your OpenAI key to get started. While llm-term offers an option for gpt-4o, it works great with gpt-4o-mini. So it's not costly. I appreciate any feedback or suggestions.
2024 · github.com
- 17CM
Hey HN, I've been building AutoAgents, an AI agent framework in Rust. Today I'm sharing a feature I haven't seen done well elsewhere: composable middleware layers for LLM inference pipelines. The problem Every agent framework lets you swap LLM providers. Almost none of them give you a structured way to enforce safety, caching, or data sanitization in the inference path itself. You end up with guardrails as application-level if-statements, caching bolted on as a separate service, and PII handling as a "we'll add it later" TODO that never ships. This gets worse with local models. Cloud APIs…
Mar 2026 · github.com
- 18
- 19TI
I'm an "ideas person" who messes around with AI on a low budget. I got tired of watching my tokens vanish and context windows filling up while agents fumbled around trying to find the right thing. Agents don't flail like they used to with shell tools, but there are still weak/blind spots and back-and-forth episodes — especially when using tools in combination/sequence. So I built "tilth" today. Or rather, AI built it — every line is Opus 4.6. I spent a lot of my precious tokens getting it to "not shit" (at least several of the different vendors' AI overlords assure me it's not…
Feb 2026 · github.com
- 20SA
2025 · spiceai.org
- 21PA
Hello Hacker News! I am Bertrand from Pruna AI. With my associates, John, Rayan, and Stephan, we are fellow researchers in AI efficiency and reliability coming from TUM. We are building an optimization engine that combines compression methods (e.g. quantization, pruning, compilation, batching…) in the aim of saving compute power when running AI models. This optimization engine take one base model as input and returns a compressed model as output. It aims to help for two things: - Make various AI models faster and/or smaller for various hardware (because they can require significant…
2024
- 22TF
I’d originally launched my app: Private LLM[1][2] on HN around 10 months ago, with a single RedPajama Chat 3B model. The app has come a long way since then. About a month ago, I added support for 4-bit OmniQuant quantized Mixtral 8x7B Instruct model, and it seems to outperform Q4 models at inference speed and Q8 models at text generation quality, while consuming only about 24GB of RAM[3] at 8k context length. The trick is: a) to use a better quantization algorithm and b) to use unquantized embeddings and the MoE gates (the overhead is quite small). Other notable features include many more…
2024
- 23R5
Hi HN, I built OpenGraviton, an open-source AI inference engine that pushes the limits of running extremely large LLMs on consumer hardware. By combining 1.58-bit ternary quantization, dynamic sparsity with Top-K pruning and MoE routing, and mmap-based layer streaming, OpenGraviton can run models far larger than your system RAM—even on a Mac Mini. Early benchmarks: TinyLlama-1.1B drops from ~2GB (FP16) to ~0.24GB with ternary quantization. At 140B scale, models that normally require ~280GB fit within ~35GB packed. Optimized for Apple Silicon with Metal + C++ tensor unpacking, plus…
Mar 2026 · github.com
- 24PC
Hi HN, We needed a simple way to connect to the top AI models to experiment, prototype and evaluate them. Main features: - Connect to top LLMs in few lines of code (currenly OpenAI, Anthropic and AI21 are supported) - Response meta includes tokens processed, cost and latency standardized across the models - Multi-model support: Get completitions from different models at the same time - LLM benchmark: Eevaluate models on quality, speed and cost The benchmark uses predefine questions to test AI reasoning abilities across a range of "hard" queries. The outputs are then automatically evaulauted…
2023 · github.com
Ranked by how close each launch is in meaning, then by votes. Refine with a description →