LLM Reference

LLM Reference is a decision-support directory and trendsetting intelligence hub built for engineers and technology leaders who need to navigate the exploding large language model ecosystem with precision and speed. In a landscape where new models, price cuts, and benchmark updates arrive weekly, this platform serves as your early-adopter compass, tracking over 1,800 language models from 140 providers and 247 research labs. The core value proposition is refreshingly simple: stop wasting hours hunting through scattered sources and start shipping with confidence. Whether you are building a cutting-edge coding assistant, an agentic workflow, a sophisticated writing tool, or a complex research pipeline, LLM Reference gives you a single, trustworthy place to compare models side-by-side, identify the cheapest frontier output pricing, and browse curated editors' picks for specific tasks like coding, agents, writing, research, image generation, and video creation. The site is designed for fast triage, enabling you to quickly identify the right model for your job, determine the most cost-effective provider, and get back to building. With a real-time Pulse feed highlighting what changed this week, including new models, price cuts, and benchmark refreshes, LLM Reference keeps you informed without the noise. It is built by the Data Advantage project and updated daily, making it an essential resource for anyone who needs to stay current with the future of AI. This is not just a directory; it is your curated lens into what is next.

Curated Editors' Picks for Every Task

LLM Reference goes beyond raw data by offering hand-selected editors' picks for specific use cases like coding, agents, writing, research, image generation, and video creation. These picks are based on rigorous analysis of the latest benchmarks, real-world performance, and cost efficiency. For example, Claude Fable 5 is the top pick for coding with an 80.3% SWE-bench Pro score, while Veo 3.1 leads for video with 30-second 4K clips. This feature eliminates guesswork and provides a fashion-forward starting point for any project.

Comprehensive and Up-to-Date Model Directory

The platform tracks over 1,800 language models from 140 providers and 247 labs, with data refreshed weekly. This includes new releases, verified price changes, and benchmark updates. The directory is searchable and filterable by task, provider, or benchmark score, allowing you to quickly find the exact model you need. With 177 new models, 53 price cuts, and 368 benchmark refreshes reviewed in a single week, this feature ensures you are always working with the most current information in a fast-moving market.

Side-by-Side Model Comparison Tool

LLM Reference includes a powerful comparison tool that lets you evaluate two models side-by-side across key metrics like benchmark scores, pricing, and context length. This feature is essential for making informed decisions when choosing between frontier models like Claude Opus 4.8 and GPT-5.5. The comparison is presented in a clean, visual format that highlights strengths and weaknesses, enabling you to quickly identify the best model for your specific requirements without sifting through fragmented sources.

Real-Time Pulse Feed and Changelog

The Pulse feed is a living dashboard that captures everything that changed in the model market this week. It surfaces new models, verified price reductions, and benchmark refreshes in a single, scannable view. The Changelog provides a historical record of updates, making it easy to track trends over time. This feature is designed for early adopters who need to stay ahead of the curve, ensuring you never miss a critical update that could impact your development timeline or budget.

Selecting the Best Model for a Coding Assistant

A lead engineer building an AI-powered coding assistant needs a model that excels at code generation, debugging, and understanding complex software architectures. Using LLM Reference, they can navigate to the Developers section, view the Coding editors' picks, and see that Claude Fable 5 leads with an 80.3% SWE-bench Pro score. They can then compare it against GPT-5.5 and DeepSeek V4 Pro to evaluate cost, context length, and benchmark performance, ensuring they ship with the most capable and cost-effective model.

Optimizing Costs for Agentic Workflows

A technology leader deploying agentic workflows across customer support automation needs a model that balances performance with cost. By using the Providers section and the Frontier Pricing signal, they can identify that Hunyuan HY3 Preview via Tencent Cloud TI Platform offers the cheapest frontier output at $0.260 per 1M tokens. They can then cross-reference this with the Agents editors' pick, Claude Sonnet 4.6, to see if the cost savings justify a trade-off in performance for their specific use case.

Researching the Latest Video Generation Models

A creative director exploring video generation for a marketing campaign wants the best quality output available. They can go to the Creatives section, view the Video editors' picks, and see that Veo 3.1 is the top choice, offering 30-second clips with native audio and up to 4K resolution through Vertex AI. They can then compare it against Runway Gen-4.5 and Wan 2.7 using the comparison tool, checking benchmark scores and pricing to make a trendsetting decision for their project.

Staying Informed on Weekly Model Market Changes

An AI researcher needs to stay current with the rapid pace of model releases and price changes without spending hours reading scattered blog posts. They can bookmark the Pulse feed and check it weekly to see a curated list of 177 new models, 53 price cuts, and 368 benchmark refreshes. This single source of truth allows them to quickly identify new entrants like DiffusionGemma 26B A4B IT and understand shifts in the competitive landscape, ensuring their research is always based on the latest data.

How often is the model directory updated?

The model directory is updated daily, with a comprehensive refresh every week. This weekly refresh includes new model releases, verified price changes from providers, and updates to benchmark scores across major suites. The Pulse feed specifically highlights what changed in the last seven days, including the number of new models, price cuts, and benchmark refreshes.

What criteria are used for the Editors' Picks?

Editors' Picks are based on a combination of benchmark performance, real-world task suitability, cost efficiency, and expert analysis. For example, the coding pick prioritizes scores like SWE-bench Pro and OSWorld-Verified, while the writing pick emphasizes Chatbot Arena ELO and qualitative assessments of output quality. Each pick is researched and updated regularly to reflect the latest model capabilities and market shifts.

Can I compare models from different providers?

Yes, the Compare tool allows you to select any two models from the directory and view them side-by-side. You can compare benchmark scores, pricing per token, context length, and other key attributes. This is particularly useful for evaluating models from different providers, such as comparing Claude Opus 4.8 from Anthropic against GPT-5.5 from OpenAI to determine which offers better value for your specific task.

Is there a way to track pricing changes over time?

Yes, the Pulse feed and the Changelog track verified price cuts and changes from providers. You can see a weekly summary of price reductions, and the Changelog provides a historical record. Additionally, the Frontier Pricing signal on the homepage shows the cheapest output cost per 1M tokens for top-tier models, giving you a real-time snapshot of the most cost-effective options available.