Currently available models:

There are two kinds of models: cloud models, served through OpenRouter, and on-prem models, which run on our own server (free).

On-prem (sph-onprem/*)

These run locally on our server. Requests are not retained and stay within the ETH network; they never go to the public internet.

  1. Qwen 3.8 27B: sph-onprem/qwen3.8-27b
  2. Nemotron 3.5 Lightning: sph-onprem/nemotron-3.5-lightning

Cloud (openrouter/*)

Commercial models. Requests are routed upstream through OpenRouter to the model's provider.

  1. DeepSeek V4 Flash: openrouter/deepseek/deepseek-v4-flash
  2. Google Gemini 2.5 Flash Image: openrouter/google/gemini-2.5-flash-image (image generation)
  3. Google Gemini 3.1 Flash Lite: openrouter/google/gemini-3.1-flash-lite
  4. Google Gemini 3.5 Flash Lite: openrouter/google/gemini-3.5-flash-lite
  5. Google Gemini 3.7 Flash: openrouter/google/gemini-3.7-flash
  6. Google Gemini 3.8 Flash: openrouter/google/gemini-3.8-flash
  7. Meta Llama 3.3 70B Instruct: openrouter/meta-llama/llama-3.3-70b-instruct
  8. Mistral Voxtral Small 24B 2507: openrouter/mistralai/voxtral-small-24b-2507
  9. Moonshot AI Kimi K3: openrouter/moonshotai/kimi-k3
  10. OpenAI GPT-4.1 Mini: openrouter/openai/gpt-4.1-mini
  11. OpenAI GPT-6 Luna: openrouter/openai/gpt-6-luna
  12. OpenAI GPT-6 Sol: openrouter/openai/gpt-6-sol