How do I get a Ollama API key?

Create an account at https://ollama.com, open the API keys page, and generate a key. Then paste it into the Dracon key manager, where it is stored encrypted and used browser-direct.

Ollama API key setup

Run open-source LLMs locally or in the cloud. Free local inference on your own hardware. Optional cloud plans for larger models with datacenter-grade GPUs.

Run it locally

Install Ollama on your machine and start the local server (default http://localhost:11434).
Open the Dracon Chat. Dracon detects the running server and lists its models in the picker. No API key needed.
Tip: HTTPS pages can block http://localhost as mixed content. Run a tunnel or serve Ollama over HTTPS if the chat can't reach it.

Free tier

Yes. Free local inference. Cloud pay-as-you-go is $0 with 3 concurrent requests and no subscription (ollama.com/pricing, checked 2026-10-27). no credit card required

Rate limits: Local: unlimited. Cloud PAYG $0: 3 concurrent requests.

Paid plans compared

Every tier Ollama publishes, with the price and the limit as the vendor states them, last checked 2026-10-27. Confirm with the provider before you buy. You pay them directly, no Dracon markup.

Ollama subscription plans: price, usage limit and what each tier includes.
Plan	Price	Usage limit	What you get
Free	no monthly fee	1 cloud model at a time; Light usage; PAYG $0, Pro $20/mo ($200/yr), Max $100/mo, Team $500/mo confirmed on ollama.com/cloud (checked 2026-10-27); board carries Free/Pro/Max.	Local inference Cloud models
Pro	$20/mo	3 cloud models at a time; 50x more than Free; PAYG $0, Pro $20/mo ($200/yr), Max $100/mo, Team $500/mo confirmed on ollama.com/cloud (checked 2026-10-27); board carries Free/Pro/Max.	Larger cloud models Upload private models
Max	$100/mo	10 cloud models at a time; 5x more than Pro; PAYG $0, Pro $20/mo ($200/yr), Max $100/mo, Team $500/mo confirmed on ollama.com/cloud (checked 2026-10-27); board carries Free/Pro/Max.	Maximum cloud usage Priority access

Official plan source: ollama.com/cloud

Connection details

Base URL: http://localhost:11434/v1
API shape: OpenAI-compatible
Env var: OLLAMA_API_KEY
Models: llama3.3, qwen2.5
Model lineup: Llama 3.3, Qwen 2.5, DeepSeek V3, Mistral, Gemma 2, Phi-4

Gotchas

Local Ollama: no key needed; runs on your machine. HTTPS pages may block http://localhost (mixed content): serve over http or use a tunnel.

Our verdicts on this kind of provider

Independent comparisons we keep current as pricing and free tiers shift. A guide is setup; a verdict is whether it's worth setting up.

Add your Ollama key Add this provider to a fallback list Compare provider plans