Alternatives
Products that do what Makes local LLMs faster and more reliable by optimizing for your device does
Time to first token is 39% faster Agent wall times decrease by 46% No swaps Tracks your resource usage in real-time and adjusts how the model runs so that it works perfectly on your device. Implements KV cache sizing, prefix caching, live RAM pressure management, context trimming, KV quantization, and more. Built a ton of features
- 1IV
The video demo runs a 7b Model on a normal gaming GPU. I think it already works quite well (accounting for the limited hardware power). :)
2024 · github.com
- 2KR
I discovered that in LLM inference, keys and values in the KV cache have very different quantization sensitivities. Keys need higher precision than values to maintain quality. I patched llama.cpp to enable different bit-widths for keys vs. values on Apple Silicon. The results are surprising: - K8V4 (8-bit keys, 4-bit values): 59% memory reduction with only 0.86% perplexity loss - K4V8 (4-bit keys, 8-bit values): 59% memory reduction but 6.06% perplexity loss - The configurations use the same number of bits, but K8V4 is 7× better for quality This means you can run LLMs with 2-3× longer…
2025 · 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
- 4
- 5
- 6UD
Hey HN! I’m the founder of Unify, and we’ve just released our Model Hub, which provides a collection of LLM endpoints with live runtime benchmarks all plotted across time: https://unify.ai/hub A key finding is that static tabular runtime benchmarks for LLMs simply do not work. It’s necessary to take a time-series perspective, and plot the variations through time. We currently have 21 models provided by: Anyscale, Perplexity AI, Replicate, Together AI, OctoAI, Mistral AI and OpenAI, with more on the roadmap. We test across different regions (Asia, US, Europe), with varied…
2024
- 7IB
Built a simple web app that tells you which open-source LLMs will work on your hardware. It auto-detects your specs, shows compatible models from Hugging Face, gives realistic performance estimates (tokens/sec), and recommends quantization settings. You can also manually input specs to see "what if I upgraded my RAM?" Made this after wasting time downloading giant models only to find they crawled on my hardware. Hope it saves you some frustration!
2025 · caniusellm.com
- 8KP
I thought it'd be interesting to use Linux PSI (Pressure Stall Information) for an LLM runtime to trim the KV cache. This is mainly useful imo for edge devices like the Jetson Orin super nano kit which have unified memory. I haven't benched much, but plan to do so more over time and see if I can make a real use of it as I run local LLMs. Let me know if it makes sense :P (I of course vibed this idea)
Jun 2026 · github.com
- 9NT
With the latest launch from Google I've added support for Gemma 3 270M, the speed for local LLM to TTS token time is incredible! This is an heavy obvious work in progress - any contributions or tips would be welcome. The idea is to have a fast moving edge model playground, and maybe have some utility (like the e reader) on the side.
2025 · github.com
- 10TF
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
- 11RM
Apr 2026 · 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
- 13AL
Dec 2025 · github.com
- 14RC
Hello HN! We're building a caching solution for LLMs (ChatGPT, Claude). By combining cutting-edge approaches, such as edge computing, prompt compression, vectorization, and others - it can reduce your AI bills by up to 10x and significantly lower response times. Key Features: - cost efficiency: our system stores frequent queries, reducing the number of upstream (paid) API calls - fast responses: with various nodes globally, we reduce latency by serving data from the nearest location - scalability: designed to handle increasing loads and data sizes without degrading performance. The cache…
2024 · edgematic.dev
- 15LC
The standard AI energy debate compares server-side LLM inference to a server-side Google query. I think this misses most of what actually happens on a mobile device during a real search session. I built a parametric model of the full end-to-end mobile search session: 4G/5G radio energy, SoC rendering cost for a 2.5MB page, programmatic advertising RTB auctions running in the background, and network transmission costs for both sides. Then compared it to an equivalent LLM session. Main finding across 10,000 Monte Carlo draws: on mobile, a standard LLM session uses on average 5.4x less…
Apr 2026 · dupr.at
- 165L
We've built InferX, a specialized runtime environment that fundamentally changes how LLMs are served. The core problem we solve is the latency bottleneck in AI inference, especially with large models. Current systems waste resources or suffer from painfully slow cold starts. InferX's AI-native architecture, with its "snapshot" technology, enables: * *Sub-2s cold starts:* Spin up models instantly. * *High density:* Serve more LLMs on the same GPUs. * *Optimal efficiency:* Maximize GPU utilization. This isn't just another API; it's a new execution layer designed from the ground up for the…
2025 · github.com
- 17BA
Hi HN, Erik here. Today we launch Butter, an OpenAI-compatible API proxy that caches LLM generations and serves them deterministically on revisit. Since April, we’ve been working on this concept of “muscle memory,” or deterministic replay, for agent systems performing automations. You may recall our first post in May, launching a python package called Muscle Mem: https://news.ycombinator.com/item?id=43988381 Since then, the product has evolved entirely, now taking the form of an LLM Proxy. For a deep dive into this process, check out:…
Oct 2025 · docs.butter.dev
- 18WB
Over the past few months, as we scaled our internal AI Agents, we hit a dead end: Running LLM-generated arbitrary code in Docker is basically running naked on security due to container escape risks. But using full traditional VMs takes minutes to boot and eats too much memory to support high-density concurrency. We loved the developer experience of SaaS sandboxes on the market, but they are closed-source, expensive, and have too high a barrier to entry for self-hosting. So, our team decided to build our own. After months of grinding, using RustVMM and KVM, we built a blazing-fast,…
Apr 2026 · github.com
- 19CL
Hey HN, we’re the developers of OpenLake, an open source storage engine for offloading LLM KV caches from GPU memory into a shared tier of RAM and NVMe. We built OpenLake because KV caches are outgrowing GPU memory. A single 256K token conversation on Gemma 4 31B produces approximately 43GB of KV state, more than half the memory of an 80GB H100. The problem becomes even harder across a cluster: a prefix cached on one GPU host is unavailable when the next request lands on a different GPU, forcing the new GPU to repeat work the fleet has already completed. Once the KV cache is offloaded,…
Jul 2026 · github.com
- 20
I made this after seeing someone posit the idea online yesterday over lunch then spent some time refining it. So far it's pretty impressive IMO! Right now I am running Qwen3-30B-A3B on my 24gb unified memory m4 MacBook Pro at 50 tok/sec and this should definitely not be working for such a large model on my middling hardware. Things are detailed in the README to get up and running and DESIGN.md has details on all the choices and such made along the way.
23d ago · github.com
- 21II
2023 · github.com
- 22IB
I was overspending on GPT-4o. It was really hard to compare different models I could switch to, so I built this LLM comparison tool. It shows leaderboards, pricing, and performance data across 100+ LLMs (including all major providers and open-source models). Key features: - Live pricing comparisons - Benchmark Scores (MMLU, HumanEval, GPQA, etc.) - Context length vs cost analysis - Speed/throughput tests across providers - Quality vs price visualizations - Open source (all data verifiable) Try it out: https://llmstats.com I'd like to know your opinion :) Tech stack: Next.js,…
2025 · llm-stats.com
- 23AC
Multi-tier exact-match cache for AI agents backed by Valkey or Redis. LLM responses, tool results, and session state behind one connection. Framework adapters for LangChain, LangGraph, and Vercel AI SDK. OpenTelemetry and Prometheus built in. No modules required - works on vanilla Valkey 7+ and Redis 6.2+. Shipped v0.1.0 yesterday, v0.2.0 today with cluster mode. Streaming support coming next. Existing options locked you into one tier (LangChain = LLM only, LangGraph = state only) or one framework. This solves both. npm:…
Apr 2026
- 24AL
Raymond here from Butter.dev, an LLM response cache built as a chat-completions proxy. Today we're launching a key feature for the platform: the ability to generalize on dynamic, templated inputs. Caching at the HTTP request level has the obvious problem of generalizability. Nearly no request is identical, due to templated variables (like names) and metadata (like timestamps), so exact-match cache lookups rarely hit. We solve this at Butter by using LLMs to detect dynamic content in requests and derive their inter-relationships, allowing the cache entry to be stored as a template + variables…
Jan 2026 · blog.butter.dev
Ranked by how close each launch is in meaning, then by votes. Refine with a description →