🚀Introducing The LLM Inference Provider Leaderboard - a live-updated,... unbiased eval of API Inference products. Featuring: Abacus.AI, Anyscale, DeepInfra, Decart, Fireworks, Lepton AI, Together AI, Perplexity, Replicate, as well as OpenAI and Anthropic models For each provider's Mixtral-8x7B and Llama-2-70B-Chat public endpoint, we benchmark cost, rate limit, P50 & P90 of throughput & TTFT, and average daily collections overtime for long term tracking. At Martian, we route each API request to the best LLM to reduce cost, reduce latency, and get the best performance. So finding the best providers is an important problem for us. We found that there's a > 5x cost difference, > 6x throughput variation, and even larger rate limit discrepancies among providers! Choosing between different LLMs is only part of the equation -- the selection of different inference endpoints is also crucial to get the best performance for your use case. See highlights of provider performance in🧵👇show more

Monitor and control your AI spend on every provider... on Our early users save 40% on average. Every week, the price-intelligence-latency frontier shifts, and we expect this trend to continue. Tradeoffs between latency, reasoning, cost, service tier, open source and closed source models are shifting constantly. Router sends every request to the model that's actually best for the task and helps you control what tokens you buy. We benchmark it against real work: ~40% lower cost for the same outputs. Today we're opening it to everyone. Two lines of code or just change your base URL. No Ramp account needed. Free through 2026, first $26 on us. Get an API key today atshow more

⭐The Year of Inference is here. Featherless is now... an official inference provider on Hugging Face, unlocking 6,700+ LLMs for anyone to run, eval, and deploy instantly. It all starts with accessibility. From DeepSeek to Mistral, LLaMA to Qwen — powerful LLMs are one click away. We believe the future of AI is shaped by the long tail: personalized, specialized models tuned to real people’s needs. To get there, inference must be open, affordable, and usable by all. Whether you're fine-tuning, prototyping, or scaling a product, this moment is for you. 🫱🏻‍🫲🏻Let’s make inference the easiest part of building with AI. 📢 Share this so more builders know what’s now possible. Excited to be partnering with clem 🤗 Julien Chaumond Vaibhav (VB) Srivastav Simon Brandeis & Hugging Face team to take this to the next level!show more

Introducing LLaMa Chat: The fastest way to chat with... AI at Meta/Meta's Llama2! Try it here: feedback appreciated! This is our first step towards building a blazing-fast LLM inference completely in-house, for everyone to try. What's next? We'll be hooking up the bigger LLAMAs and eventually deploying our own in-house LLMs. All of this in service of building the best and fastest answer engine. Come join us, we're hiring!show more

"The best organizations won't manage token budgets manually. Instead,... they'll rely on orchestration layers that automatically route workloads to the right models based on performance, cost, and use case. Keeping track of which model is best for coding, finance, research, or support will become too complex for humans to manage directly, so intelligent routing will become a core part of enterprise AI infrastructure." Aravind Srinivas How will the best orgs of the future manage token budgeting Michael Mignano Nikesh Arora Ryan Petersen mmurphshow more

Something NVIDIA & Google do better than anyone else... is software-hardware-system co-design, and not just optimizing hardware for current model architectures, but predicting future ones. Back in early 2022, when NVIDIA started the design process for NVL72, MoE (Mixture of Experts) models were not yet the standard, and dense models were still dominant for frontier models. However, NVIDIA's strong software-hardware co-design culture enabled them to make a calculated bet that MoEs were the future, and they built NVL72 specifically for best MoE performance per TCO (Total Cost of Ownership). Furthermore, back in 2022, disaggregated prefill and wide expert parallelism (wideEP) MoE inference optimizations hadn't been invented yet, but it turns out that these MoE inference optimizations work best on large-scale systems like NVL72. While most other AI chip companies' in-house AI labs focus on training small 5B models that mainly use data parallelism, NVIDIA and Google's in-house AI labs continuously push the boundaries of model architecture and training recipes, such as NVFP4 training. Just like Super Idol & IShowSpeed, there must be a strong partnership between software engineers and hardware engineers to deliver the best systems that maximize performance per TCO.show more

Season 2 of Fableborne went live this morning and... we are now trending in gaming! Hundreds of thousands of raids already complete! For 2 years we waited and waited, patiently building, restraining ourselves from all the noise and short term distractions focsuing just on building the best game to use as the base of what is to come. As a founder, working with this team each day, its easy for me to be unrealistically bullish. Its so satisfying to see the rest of web3 now seeing why.show more

$ZKHIVE X $NMT 🤝 We are excited to announce... our major partnership with one of the leading AI projects in the space - NetMind.AI ! $NMT is an award winning AI powerhouse that has quickly become one of the most recognized decentralized AI platforms in the space, thanks to their wide range of offerings, from cost effective compute farms to elite AI consultancy services and AI chatbots. As part of our long term partnership, $ZKHIVE and NMT are going to build security AI models together 🤖🐝 #zkHive will also receive an access to $NMT’s compute farm as well as AI consultancy from $NMT’s team of experts, and $NMT will get access to all of #zkHive’s products and APIs! This partnership will improve zkHive’s AI capabilities and introduce thousands of new users to $ZKHIVE ecosystem!show more

Day 11/90 of Inference Engineering How does vLLM work... and how is it used in production? Before we discuss how vLLM works internally, it helps to understand what vLLM is. At a high level, vLLM is an inference engine that is designed to serve LLMs to thousands of concurrent users efficiently while managing scarce compute and memory. The goal for vLLM is to maximize throughput and minimize latency; optimizing for the best inference economics and experience for end users. With every request from the end user, it eventually ends up in the engine core, gets scheduled alongside other requests from other concurrent users, executes on the GPU, and updates the KV cache with the new key and value vectors, and streams the tokens back to the user. The Scheduler decides what requests should execute next while continuously batching requests together to maximize GPU utilization. Continuous batching is an inference optimization that allows new requests to join a running batch as other requests finish generating tokens. This helps with keeping the GPU utilization high instead of letting it sit idle waiting for an entire batch to complete generating. After the scheduler dispatches the selected batch to the Model Executor, the Model Executor prepares the tensors and metadata required for inference, retrieves each request’s block table from KV Cache Manager, launches the optimized transformer forward pass on the GPU, computes the logits, updates the KV cache with the new key and value vectors, and finally returns the results for sampling and streaming. The KV Cache Manager uses the PagedAttention memory layout to allocate fixed-size cache blocks on demand and maintains a Free Block Queue on the CPU that tracks which blocks in the GPU’s Paged KV Cache are currently free. When a request needs additional KV cache space, the KV Cache manager takes a free block from the queue and assigns it to that request, thus avoiding an expensive search through GPU memory for available cache blocks. All of these components form the core of vLLM’s inference engine. The Scheduler determines what requests are executed, the Model Executor determines how those requests are executed, the KV Cache Manager determines where each request’s KV cache lives using the PagedAttention Memory Layout. This architecture enables vLLM to serve thousands of concurrent requests with high throughput, low latency, and efficient GPU memory utilization. Heres a little animation that visualizes everything! - I've also completed the forward pass for my mnist.c project. I had a nice chat with shrey birmiwal, such a knowledgeable guy. Excited to learn more about vLLM and implement a tiny-vLLM one day.show more

At Dreamforce, the exploration of the possibilities of AI... is incredible. At the same time, it's important that we understand that policies never move at the same speed as technology. So having a grasp over what is happening and how the negative aspects of AI don't proliferate in a big way, to minimize collateral damage is most important. I hope the companies and governments across the world wake up to this and ensure the best of AI possibilities will reach the people, as this is a phenomenal enabler for humanity like never before. -Sg Marc Benioff Dreamforce Salesforce #DF24show more

We’re excited to introduce ShinkaEvolve: An open-source framework that... evolves programs for scientific discovery with unprecedented sample-efficiency. Blog: Code: Like AlphaEvolve and its variants, our framework leverages LLMs to find state-of-the-art solutions to complex problems, but using orders of magnitude fewer resources! Many evolutionary AI systems are powerful but act like brute-force engines, burning thousands of samples to find good solutions. This makes discovery slow and expensive. We took inspiration from the efficiency of nature. ‘Shinka’ (進化) is Japanese for evolution, and we designed our system to be just as resourceful. On the classic circle packing optimization problem, ShinkaEvolve discovered a new state-of-the-art solution using only 150 samples. This is a big leap in efficiency compared to previous methods that required thousands of evaluations. We applied ShinkaEvolve to a diverse set of hard problems with real-world applications: 1/ AIME Math Reasoning: It evolved sophisticated agentic scaffolds that significantly outperform strong baselines, discovering an entire Pareto frontier of solutions trading performance for efficiency. 2/ Competitive Programming: On ALE-Bench (a benchmark for NP-Hard optimization problems), ShinkaEvolve took the best existing agent's solutions and improved them, turning a 5th place solution on one task into a 2nd place leaderboard rank in a competitive programming competition. 3/ LLM Training: We even turned ShinkaEvolve inward to improve LLMs themselves. It tackled the open challenge of designing load balancing losses for Mixture-of-Experts (MoE) models. It discovered a novel loss function that leads to better expert specialization and consistently improves model performance and perplexity. ShinkaEvolve achieves its remarkable sample-efficiency through three key innovations that work together: (1) an adaptive parent sampling strategy to balance exploration and exploitation, (2) novelty-based rejection filtering to avoid redundant work, and (3) a bandit-based LLM ensemble that dynamically picks the best model for the job. By making ShinkaEvolve open-source and highly sample-efficient, our goal is to democratize access to advanced, open-ended discovery tools. Our vision for ShinkaEvolve is to be an easy-to-use companion tool to help scientists and engineers with their daily work. We believe that building more efficient, nature-inspired systems is key to unlocking the future of AI-driven scientific research. We are excited to see what the community builds with it! Learn more in our technical report:show more

Moonshot AI is casually giving developers free daily access... to Kimi K3 😳 no subscription no upfront payment just sign in and start using one of the largest open AI models available what you get for $0: - Kimi K3 with 2.8T parameters - 1M token context window - strong coding and reasoning performance - native vision capabilities - free daily credits that refresh automatically why this is worth checking: > access a frontier model without paying API fees > long context for large codebases and documents > works on web, desktop, mobile, and CLI getting started takes less than 2 minutes: 1. go to 2. create a free account 3. Kimi K3 is available as the default model 4. start chatting or coding with your daily free credits bonus: Moonshot Together lets you invite friends for a chance to earn 3, 7, 15, 30, or even 365 days of Kimi Membership through its rewards program benchmark highlights: > 2.8T parameter MoE model > 1M context window > strong performance across coding, browsing, and reasoning benchmarks important: free credits reset daily, rate limits apply on the free tier, and the open-weight release is expected on July 27 A simple way to try one of the latest frontier AI models without paying for API accessshow more

S5 🚨: Anthropic is one of the most well-positioned... AI labs to disrupt consumer market this week, and all stars are aligning for them. - Super Bowl is the single biggest marketing event of the year - Sonnet 5, as a model name, is the best background for an advertisement - Cowork is an overpowered consumer tool that almost no one knows about - Sonnet 5 performance is at the next level Some S5 (non-thinking) generations 👀 h/t Леонидshow more

I’m incredibly proud to share that OpenAI chose Brex... to power their global spend and financial operations. When you're building at the frontier of AI and scaling global teams and infrastructure at an unprecedented pace like OpenAI is, Finance can't be the thing that slows you down. You need spend visibility the moment it happens, controls that enforce themselves, and agentic workflows that eliminate the manual work so your team stays focused on driving the business forward. We were so impressed by OpenAI’s rigor in evaluating every solution in the market, and whether they align to the agentic future OpenAI is building. Their decision to run on Brex is a huge testament to our AI roadmap and vision for the future of Finance. We started Brex around a simple idea: companies shouldn't have to choose between speed and control. There's no company in the world where that tradeoff matters more than OpenAI. We are honored to support them as they build the future. The best AI companies in the world, including OpenAI, Anthropic, Cursor, Vercel, Granola, Sierra, and Mercor choose Brex over every alternative for that exact reason. If you want to understand who’s truly building the future of AI in Finance, follow the customers you admire the most – not the hype.show more

There is only one reason to share truth. And... no it has nothing to do with money- and everything to do about helping God move. Peace is the prize- a better world for everyone is the prize. We make SACRIFICES for things we want most in this world. Christ made the ultimate sacrifice If He can do that for us FREE of charge- we owe it to each other to work TOGETHER to make the world a better place- despite the toll and sacrifice. A level playing field- where we can all work hard and do our best to succeed- that is the payout. A better world for all We’re not here for just ourselves- otherwise the second commandment wouldn’t be like the first- love your neighbor as yourself. This world gets better the sooner we all wake up. Information is free- we collect and share it for free in hopes that it keeps spreading. Only working TOGETHER can we win. We are meant to move these messages for God so that humanity can wake up- this plan is designed that way- working together The Lord is ONE. We are bound as a family by Our Father in Heaven. We are apart of Him and He is apart of all of us. We are billions of little pieces of the Lords gift of life. We’ve been separated and pushed apart because it makes us weak. We were designed to be powerful by the love we draw from Creator and push through creation. We are ONE with Our Holy Father 🙏💗 Be EXCELLENT to each other frensshow more

A good technical LLM interview question: Your LLM chatbot... takes 12s before it generates the first token, and the users are complaining. So you move the model onto a GPU with 3x the computing power. The time to first token barely improves. Why did this happen? (answer below) Latency in an LLM app is a placement problem disguised as a model problem. If you profile the 12 seconds, the model's prefill itself may only account for around 1.5 seconds of it. So halving the prefill step saves just 750ms out of 12000, which is under 7%. The rest is spread across stages that never touch the GPU. The request first travels to whatever region the app runs in, and a cross-continent round trip could cost over a second before any code executes. Then the request handler starts. On a container-based serverless platform under load, this adds several seconds of cold start, paid before auth, rate limiting, or prompt assembly even begins. Retrieval adds its own hop, and the response streams back across the same distance. Optimizing a stage that was already fast cannot alter the latency that's majorly affected by other stages. Those other stages are slow for a structural reason. An LLM app runs two workloads that want opposite machines. - The request path is short, spiky, and needs to sit close to users - Inference is long-running, GPU-bound, and billed hourly, whether requests arrive or not. So the actual decision is not which model to run, but where each of these two workloads runs. There are three options, each with its own tradeoffs: > A dedicated GPU box removes inference cold starts, but it bills around the clock and lives in one location, so distant users wait out the round trip on every request > Container-based serverless scales to zero, but the request path pays a cold start, and most of these platforms have no GPU behind them. > Edge runtimes start in under a millisecond, because a WebAssembly module carries no OS or container image to boot. They handle the request path well and cannot hold a model. So the answer is not to pick one, but to split the app across two of them. The request path runs close to users, and inference runs on a dedicated GPU it calls into. That also explains the failed upgrade. More compute made a stage that was already fast faster, and left the 10.5 seconds around it untouched. To actually learn how it's done in practice, Akamai's GitHub has a reference implementation for each half. - vllm-on-lke serves Qwen2.5-7B-Instruct behind an OpenAI-compatible endpoint on one RTX 4000 Ada GPU in Linode Kubernetes Engine, with Terraform creating the cluster, both firewalls, and the GPU operator in one apply. - akamai-functions-llm-chatbot covers the front, where a WebAssembly API checks a KV cache and only calls the GPU-backed instance on a miss. Both are available on Akamai’s new Developer Hub, alongside their tutorials and code samples. It also links to Edge Case, their Discord, where four developer advocates architect and deploy a production app live every other Wednesday. If you create a new Akamai Cloud account, you can also get $300 in credits for joining. Join here: That said, this post treats generation as a single 1.5s block, but that block has its own structure, and knowing it well tells you whether a model is slow to start or slow to stream. I wrote a first-principles walkthrough of it, covering the prefill and decode split, KV caching, and where the time actually goes inside each one. Read it below. Thanks to Akamai Cloud for partnering today!show more

Unpopular opinion: Most agent evals are theatre. You run... them once before the deployment. It'll take 800ms+ as another LLM would be judging your LLM. Most annoying part - no one tells where in the chain things went wrong. I wasted a lot of time in this loop. And then I came across Future AGI bringing 5 different tools under one umbrella, best part - the platform is completely open source. They open sourced their entire platform and the eval layer is noticeably different. It is multimodal - works on everything text, image, audio, pdf. Not an LLM-as-judge adding latency but an agent with memory and tools. The biggest win are learned classifiers trained on actual production failure patterns to run evals at low cost. It also runs across the full reasoning chain, not just the final response. Check out → Try it here →show more

We are entering an extremely exciting era for open-weight... models. Kimi K2.6 now feels like a top agentic model. I took it for a spin via Fireworks AI fast inference APIs. Kimi K2.6 has impressive agentic capabilities, design skills, and the ability to synthesize large amounts of information. I built a little Skill that produces survey papers on any AI research topic you want. (see example in the clip) You can use the skill to tell your agent to generate a survey on whatever topic and watch it go to work. The artifact was fully generated by Kimi.ai's Kimi K2.6. It's cheap and fast. Next step for me is to explore ways to continue integrating the capabilities of these models on use cases like automating my LLM knowledge bases and augmenting my agent memory capabilities. Stay tuned for more.show more

LayerAI AI Agent Manifesto is Live: The Path Forward... 🧬 We've made it easy for ecosystem veterans & newcomers to get excited about the market & product opportunity we're tackling next: AI Agent Infrastructure. We're building an AI-powered agent platform where people can deploy, market, and succeed with this new token subcategory. At the heart of this transformation lies a challenge: primitive & so far limited tech & AI capabilities of incumbent platforms. We believe that LayerAI is equipped to rise as the new leading infrastructure provider for this market. LayerAI has already demonstrated market validation for AI Agents and looks to build on what we believe is the very start of this category in web3. 👉 Explore now:show more

This was an inflection point for me yesterday when... I became a part of Paramhansa Yoganada Charitable Trust in Vrindavan - an organisation working for the welfare of the widows Visiting there made me realise that there is no definition of “enough” for each one of us in life. Whatever we have is a dream for many others. Being content is the primary goal for me at least Having a good house to live and basic needs fulfilled, will try my best to parlay my trading gains for the lives of others. It had been a life-changing day for me…show more

AI has had exactly two scaling axes that worked... so far, and the second one is starting to look finite too the first one was pretraining: with scaling parameters and data, we got world knowledge (i.e. ChatGPT had read enough to know things), but it started saturating a while ago the second one was RL, and people had been doing RL the whole time before that: RLHF is RL but it never scaled far because it was trying to control the exact output, which tokens come out, how the text reads, but you can only push that so far before you’re just polishing RLVR dropped that constraint: giving the model a task, then checking whether the final answer is right, and ignoring everything in between -- so the model does whatever it wants in the middle and only the endpoint gets graded, and that’s much closer to actual RL and it’s what bought us planning and reasoning (arguably, tool use sits around 2.5 on this list -- while useful, it's not a different kind of thing) so one axis gave knowledge, the other gave reasoning, and both of them are one model working alone the next axis is how many models you can get working on the same problem, which is a different kind of axis than the previous two we know that multi-agent RL has always been the harder problem: I spent years in that literature and the gap between single-agent and multi-agent is definitely not incremental -- it’s a whole different class of difficulty! which is also why the derivatives are steep at the start, nobody has picked the easy wins yet... and the thing that gates this multi-agent coordination is communication: models can only coordinate as well as they can exchange information, and right now they do that by writing sentences to each other imagine what could we possibly achieve if we properly open that third axis development by letting models to exchange information in their native "language" without loosing any computational data that they produce during inferenceshow more