Support Open Source

Esper 3.1 now available for Qwen3.6-35B-A3B and Qwen3.6-27B, plus other small models!

Structured, low-token medical reasoning data powering the Guardpoint model series.

Structured, low-token reasoning for medical diagnosis, management, knowledge, and understanding: available for Qwen 3.5, Qwen 3, and gpt-oss.

A creative reasoning dataset offering head-to-head comparison between DeepSeek-V3.2-Speciale and DeepSeek-V3.2.

Experimental fine-tuned models that utilize reasoning to create specific structured forms of output.

Synthetic datasets generated using reasoning models, primarily the DeepSeek-V4, Deepseek-R1 and Deepseek-V3 series.