Lista modeli Ollama

└─$ ollama ls               
NAME                    ID              SIZE      MODIFIED      
qwen2.5-coder:latest    dae161e27b0e    4.7 GB    8 minutes ago    
gpt-oss:latest          17052f91a42e    13 GB     3 weeks ago      
llama3:latest           365c0bd3c000    4.7 GB    3 weeks ago      
gemma2:latest           ff02c3702f32    5.4 GB    3 weeks ago      
deepseek-r1:latest      6995872bfe4c    5.2 GB    3 weeks ago      
starcoder:latest        847e5a7aa26f    1.8 GB    3 weeks ago    s

Żądanie cURL czatu Ollama

Teraz napiszmy proste polecenie cURL, które wywoła model *llama3:

curl http://localhost:11434/api/chat -d '{
  "model": "llama3",
  "messages": [
    { "role": "user", "content": "Give me very short answer. The capitol of Poland is?" }                       
  ]
}'

Wynik będzie podobny do:

{"model":"llama3","created_at":"2025-11-10T10:10:03.318829243Z","message":{"role":"assistant","content":"Wars"},"done":false}
{"model":"llama3","created_at":"2025-11-10T10:10:03.519302483Z","message":{"role":"assistant","content":"aw"},"done":false}
{"model":"llama3","created_at":"2025-11-10T10:10:03.703234588Z","message":{"role":"assistant","content":"!"},"done":false}
{"model":"llama3","created_at":"2025-11-10T10:10:03.883560841Z","message":{"role":"assistant","content":""},"done":true,"done_reason":"stop","total_duration":1631819270,"load_duration":103364571,"prompt_eval_count":23,"prompt_eval_duration":961241860,"eval_count":4,"eval_duration":559116162}

Czat Ollama cURL bez strumieniowania

 curl http://localhost:11434/api/chat -d '{
  "model": "llama3",
  "messages": [
    { "role": "user", "content": "Give me very short answer. The capitol of Poland is?" }
  ], 
  "stream": false
}'

Generowanie odpowiedzi Ollama

curl http://localhost:11434/api/generate -d '{
  "model": "llama3",
  "prompt": "Give me very short answer. Write a simple code in Python with hello world example." , "stream": false}
}' | jq -r ".response"