125 models 35 labs
Ai21
1 model
Aion Labs
2 models
Amazon
2 models
Anthropic
10 models
Cohere
3 models
Deepcogito
1 model
DeepSeek
4 models
Inception
1 model
Inclusionai
3 models
Kwaipilot
3 models
Liquid AI
2 models
Meta
4 models
Microsoft
1 model
MiniMax
3 models
Mistral AI
8 models
Moonshot
4 models
Nex Agi
1 model
Nvidia
4 models
OpenAI
16 models
Perplexity
1 model
Poolside AI
2 models
Qwen
16 models
Sakana
1 model
StepFun
1 model
Tencent
3 models
Upstage
1 model
Writer
1 model
xAI
7 models
Xiaomi
2 models
Z Ai
5 models

Keep exploring

How models are curated on LMRank

Every model on LMRank is listed because it is available through a public API and has at least one benchmark result we can verify. Models are not added by request from labs or sponsors. They appear when they ship, when they are measurable, and when they compete against their peers on the same tasks. No model gets preferential placement, and no lab pays to have its tools ranked higher or listed more prominently. The directory exists to give you a single, honest snapshot of what is actually out there, ranked by data rather than marketing.

Models are grouped by the lab that built them and listed alphabetically within each group so you can scan by creator or jump straight to a model you already know. Search by model name or lab to narrow the directory instantly. Rankings and pricing update as new benchmarks land and models roll out, so the leaderboard stays current without manual curation. If a model is missing or a score looks wrong, the methodology page explains how every ranking is sourced and how to flag a correction.