Compact model arena with GLM and GPT OSS commentary
Chat with every SmolLM2 instruct model.
Experiment with and compare different tokenizers
Read research updates and project news
How well can you prompt a small AI?
Every tiny LM, same eval harness, transparent benchmarks