

LLM Stats
β Text source-checked Nov 16, 2026Source-checked against the vendor's site Β· rebuilt from the vendor's own pagesProfile updated Link checked How we rate
In short
LLM Stats offers independent AI model rankings using a composite score from public benchmarks and live API metrics. It tracks 448 models, including open-weight options for self-hosting.
The vendor says that LLM Stats provides independent rankings of AI models from providers such as GPT, Claude, Gemini, Llama, and DeepSeek using a composite LLM Stats Score updated from public benchmarks and live API metrics.
The LLM Stats Score combines several categories of model evidence into one comparable number using an uncertainty-aware approach. The ranking aggregates GPQA, SWE-Bench Verified, coding-arena performance, and pricing into a single comparable score.
A dedicated Open LLM Leaderboard filters the catalog to models with publicly released weights for self-hosting or fine-tuning. Pricing data is pulled from each provider's public price list and cross-checked against billing samples through the LLM Stats proxy.
The platform tracks 448 canonical models across major labs and inference providers.
According to LLM Stats: The LLM Stats Score combines several categories of model evidence into one comparable number using an uncertainty-aware approach. Its pages list this strength: Tracks 448 canonical models across major labs and inference providers.
β The AI Tool Bible editorial team
Pros
- β Tracks 448 canonical models across major labs and inference providers.
- β Measures output speed consistently across supported provider APIs.
- β Cross-checks pricing against billing samples through the LLM Stats proxy.
- β Provides a dedicated leaderboard for open-weights models suitable for self-hosting.
Facts
- Platforms
- api (not source-checked)
- Pricing
- free (not source-checked)
- API available
- Yes (not source-checked)
- Model used
- Multi-model (not source-checked)
- Hacker News
- 5 stories Β· top 7 points Β· last Jun 19, 2026 Β· checked Nov 15, 2026
- Earliest Wayback capture
- 2024-08-11 (earliest archived capture of the site, not a launch date)
Where this tool sits
Frequently asked
- How much does LLM Stats cost?
- LLM Stats is free to use. The provided documentation does not list specific pricing tiers or additional costs for accessing the rankings or data.
- Which AI models are ranked on the platform?
- The platform tracks 448 canonical models from major labs and inference providers, including GPT, Claude, Gemini, Llama, and DeepSeek. It also features a dedicated leaderboard for models with publicly released weights.
- How is the LLM Stats Score calculated?
- The score combines GPQA, SWE-Bench Verified, coding-arena performance, and pricing into a single comparable number. It uses an uncertainty-aware approach to aggregate these categories of model evidence.
- Can I find open-source models for self-hosting?
- Yes, a dedicated Open LLM Leaderboard filters the catalog to show only models with publicly released weights, making it easier to identify options suitable for self-hosting or fine-tuning.
- Where does the pricing data come from?
- Pricing data is pulled from each provider's public price list. This information is cross-checked against billing samples through the LLM Stats proxy to ensure accuracy in the composite score.