The leading AI Infra & Developer Tools tools, models and organizations, ranked by an explainable score across popularity, momentum, credibility and ecosystem impact.
| # | Tool | Overall | Pop. | Mom. | Cred. | Eco. |
|---|---|---|---|---|---|---|
| 1 | claude-mem · MLOps Persistent Context Across Sessions for Every Agent – Captures everything your agent does during sessions, compresses it with AI, and injects relevant context back into future sessions. Works with Claude Code, OpenClaw, Codex, Gemini, Hermes, Copilot, OpenCode + More | 86 | 74 | 99 | 86 | 85 |
| 2 | LlamaFactory · MLOps Unified Efficient Fine-Tuning of 100+ LLMs & VLMs (ACL 2024) | 86 | 74 | 99 | 85 | 84 |
| 3 | airllm · Inference AirLLM 70B inference with single 4GB GPU | 85 | 74 | 99 | 83 | 82 |
| 4 | Ollama · Inference Runs open-weight models on your own machine with a one-line pull-and-run workflow, and serves them over a local API. | 79 | 63 | 76 | 92 | 92 |
| 5 | Patronus AI · Eval An evaluation and testing platform that automatically scores, monitors and stress-tests AI systems and agents. | 78 | 74 | 90 | 74 | 70 |
| 6 | Etched · Inference A startup building Sohu, a transformer-only ASIC aimed at far faster, cheaper LLM inference than GPUs. | 77 | 74 | 87 | 74 | 70 |
| 7 | OpenRouter · Inference A unified API gateway that routes requests across hundreds of LLMs from many providers. | 76 | 64 | 76 | 84 | 84 |
| 8 | Predibase · MLOps A platform for fine-tuning and serving small, specialized open-source language models efficiently. | 76 | 74 | 84 | 75 | 70 |
| 9 | d-Matrix · Inference An AI chip startup building in-memory compute accelerators for ultra-low-latency generative inference. | 75 | 74 | 78 | 76 | 70 |
| 10 | LangSmith · Eval LangSmith — Eval · AI Infra & Developer Tools. | 74 | 58 | 76 | 84 | 84 |
| 11 | NVIDIA · Inference NVIDIA — Inference · AI Infra & Developer Tools. | 72 | 74 | 38 | 98 | 94 |
| 12 | Fireworks AI · Inference A fast, low-cost inference platform for serving open-source and fine-tuned generative models via API. | 72 | 54 | 73 | 84 | 84 |
| 13 | RunPod · Inference A GPU cloud offering on-demand and serverless compute for training and deploying AI models. | 72 | 57 | 87 | 74 | 70 |
| 14 | Langfuse · Observability An open-source LLM engineering platform for tracing, evaluating and monitoring AI applications. | 72 | 54 | 73 | 84 | 84 |
| 15 | Together AI · Inference Together AI — Inference · AI Infra & Developer Tools. | 71 | 53 | 73 | 84 | 84 |
| 16 | Modal · Inference Modal — Inference · AI Infra & Developer Tools. | 71 | 55 | 84 | 75 | 70 |
| 17 | Braintrust · Eval An end-to-end platform for evaluating, testing and iterating on LLM applications in development and production. | 71 | 49 | 90 | 74 | 70 |
| 18 | LiteLLM · Inference An open-source proxy and SDK that gives a unified, OpenAI-compatible interface to 100+ LLM providers. | 71 | 50 | 90 | 74 | 70 |
| 19 | Unsloth · MLOps An open-source library that makes LLM fine-tuning dramatically faster and more memory-efficient. | 71 | 50 | 90 | 74 | 70 |
| 20 | Replicate · Inference Replicate — Inference · AI Infra & Developer Tools. | 70 | 55 | 64 | 86 | 84 |
| 21 | Fal.ai · Inference A generative-media inference platform optimized for running image, video and audio models at low latency. | 70 | 53 | 84 | 75 | 70 |
| 22 | Helicone · Observability An open-source observability platform for logging, monitoring and analyzing LLM API requests. | 70 | 46 | 90 | 74 | 70 |
| 23 | Portkey · Observability An AI gateway and observability layer that routes, caches and monitors calls across LLM providers. | 70 | 48 | 90 | 74 | 70 |
| 24 | Baseten · Inference A platform for deploying and serving machine learning models in production with autoscaling infrastructure. | 69 | 54 | 78 | 76 | 70 |
| 25 | Groq · Inference Groq — Inference · AI Infra & Developer Tools. | 68 | 58 | 55 | 87 | 84 |
| 26 | Weights & Biases · Observability Weights & Biases — Observability · AI Infra & Developer Tools. | 68 | 55 | 58 | 87 | 84 |
| 27 | CoreWeave · Inference A specialized GPU cloud provider delivering large-scale NVIDIA compute for AI training and inference. | 68 | 56 | 58 | 87 | 84 |
| 28 | Arize AI · Observability An ML and LLM observability platform for monitoring, troubleshooting and evaluating models in production. | 68 | 48 | 67 | 85 | 84 |
| 29 | Galileo · Eval An evaluation and observability platform for measuring and improving the reliability of generative AI apps. | 68 | 45 | 84 | 75 | 70 |
| 30 | Lakera · Eval An AI security company protecting LLM applications from prompt injection and other GenAI threats. | 68 | 47 | 84 | 75 | 70 |