Models
AI models you can mine on Tenseuron.
Bring a hosted provider key or run a model on your own hardware. Serve tasks, and you are paid in the same token that funds usage on that network — so what you earn can pay for your own inference. How that makes the API free.
Supported providers
Six hosted providers and two local runtimes, plus a catch-all for any OpenAI-compatible endpoint. Hosted providers need your own API key; local runtimes need no key and no API bill.
platform.deepseek.com/api_keys
platform.openai.com/api-keys
console.anthropic.com/settings/api-keys
console.groq.com/keys
console.mistral.ai/api-keys
aistudio.google.com/app/apikey
Run models locally — no API key needed
Run models locally via LM Studio server
Default models
These are the defaults each provider registers with. The model field is free text, so a miner can run anything the provider serves — these are the starting points, not the limit.
Frequently asked questions
Which AI models can I run as a Tenseuron miner?
Any model reachable through a supported provider, plus anything served by an OpenAI-compatible endpoint. Miners register with a provider — DeepSeek, OpenAI, Anthropic, Groq, Mistral or Gemini — or run models on their own hardware through Ollama or LM Studio. The model field is free text, so you are not limited to the defaults listed here.
Do I need to pay for an API key to mine?
No. Ollama and LM Studio run models on your own machine with no API bill at all. If you prefer a hosted provider you supply your own key, and it stays on your machine — it is never transmitted to Tenseuron.
Can I choose which model answers my tasks as a buyer?
Not today. You submit a task describing the job, miners choose which model to run it on, and validators score the results. Letting buyers and network creators select models is planned but not yet available.
How do I get credits without paying?
Serve tasks as a miner. Miners are paid in the network's own token, and that same token is what funds usage on that network — so you can lock what you earn and use the API without spending anything. Every wallet also gets three free tasks per network to try it before mining.
Does running a bigger model earn more?
Tasks that use tools or run in thinking mode carry a higher reward multiplier, so more capable models are worth more on that work. Miners can also declare a cost per task and a minimum margin, and tasks that would not clear that floor are never shown to them.
Start mining
One command sets up mining with whichever provider you picked. No stake is required to start.
npx @tenseuronp/cli setup