Resource Center

15/4 Weekly Update: HumanX and the Gemma 4 release

Doubleword & .txt partner to provide structured generation outputs natively through Doubleword

Introducing dw - the Doubleword CLI

27/3 Weekly Update: Doubleword CLI and OCR model release

Doubleword for OpenClaw - Your OpenClaw Agent Is Probably Burning Money It Doesn't Need To

OCR and the Bitter Lesson

20/3 Weekly Update: New Models, Free Nemotron, and Organizations

13/3 Weekly Update: Async Pipeline Generator

6/3 Weekly Update: Qwen3.5-9B + Auto Top-Up

27/2 Weekly Update: Qwen3.5-35B-A3B (Higher Quality, Lower Cost)

20/2 Weekly Update: New Qwen Models, GPT-OSS 20B & Webhooks

Scaling Curation with LLM Comparisons

LLM powered data structures: A concurrent, lock-free binary search tree

ZeroDP: Just-In-Time Weight Offloading over NVLink for Data Parallelism

Large-Scale Semantic Search Without Embeddings

Parallel Primitives for Multi-Agent Workflows

Real-Time vs Batch Inference for LLMs: Use Cases, Costs, Workflow

Behind the Stack, Ep 13 - Faster Inference: Speculative Decoding for Batched Workloads

Costco of Inference: Introducing Doubleword Batched, the Inference Provider Built for Batched Workloads

Behind the Stack Ep. 12 - Understanding Model Parallelism

Behind the Stack, Ep. 11 - How Speculative Decoding Speeds Up Language Models

Doubleword Open Sources the World’s Fastest AI Gateway

Chasing Cheap Tokens: 2x Cheaper Tokens Than H100s with Consumer Cards‍

Should GPUs make Free Trade Agreements?

Behind the Stack, Ep 10 - Batched Endpoints

What is InferenceOps? Defining the Function Behind Scalable AI

Scaling AI Requires InferenceOps, Not MLOps

Behind the Stack, Ep 9 - How to Evaluate Open Source LLMs

What the U.S. AI Action Plan Really Means for Regulated Enterprises

Behind the Stack Ep. 8 - Choosing the Right Inference Engine for Your LLM Deployment

Lightweight Prototyping or Full-Scale Ops? Ollama vs Doubleword Explained

Behind the Stack, Ep 7 - Choosing the Right Quantization for Self-Hosted LLMs

Building GenAI in Regulated Industries: A Guide to Secure, Compliant AI

Behind the Stack, Ep 6 - How to Speed up the Inference of AI Agents

Behind the Stack, Ep 5 - Making RAG Work for Multimodal Documents

Behind the Stack, Ep 4: Making Your Load Balancer LLM-Aware

GTC Europe 2025: ASAS AI & Doubleword Announce Strategic Partnership to Deliver Sovereign, Enterprise-Grade AI Solutions in Saudi Arabia and the Middle East

Doubleword doubles down on NVIDIA collaboration to give enterprises control over their AI with NVIDIA NIM microservices integration

Behind the Stack, Ep 3: How to Serve 100 Models on a Single GPU with No Cold Starts

Behind the Stack, Ep 2: How Many Users Can My GPU Serve?

Doubleword Launches Self-Hosted Inference Platform on Snowflake Marketplace

Behind the Stack, Ep 1: What Should I Be Observing in my LLM Stack?

What It Really Takes to Self-Host Your Inference Stack

Why Owning Your AI Stack Is Becoming a Strategic Advantage

AI-Powered Performance: How Digits Built Specialized Models for Accounting

Doubleword raises $12M Series A led by Dawn Capital to make self-hosted AI inference effortless for enterprises

Announcing Doubleword: New Name, Same Team, Same Mission

MLP: Attention in a Trench Coat

The Next Leap in Speculative Decoding: Inside Doubleword's Inference Engine

The End of the Centralized API Era and the Rise of the AI Sprawl

Optimising LLM Latency: Why Speed Matters In Generative AI

DeepSeek Chronicles: My Personal Take on the AI Buzz

Take Control of Your AI: Why You Should Self Host Large Language Models

Takeoff Serverless LoRA: Efficient inference at scale for fine-tuned models

Optimizing GPU Memory for LLMs: A Deep Dive into Paged Attention

Reflection on 2024 Predictions: How Did We Do?