Starting benchmark for codestral:22b at 20261015-1242...
Model information:
  Model
    architecture        llama    
    parameters          22.2B    
    context length      32768    
    embedding length    6144     
    quantization        Q4_0     

  Capabilities
    completion    
    insert        

  Parameters
    stop    "[INST]"      
    stop    "[/INST]"     
    stop    "[PREFIX]"    
    stop    "[MIDDLE]"    
    stop    "[SUFFIX]"    

  License
    # Mistral AI Non-Production License    
    ## 1. Scope and acceptance             
    ...                                    


===================================================================
Running model at 8192 context length...
===================================================================

{
  "model": "codestral:22b",
  "created_at": "2026-10-15T16:42:26.668788474Z",
  "response": " Sure, here is a simple \"Hello, World!\" program in Python:\n\n```python\nprint(\"Hello, World!\")\n```\n\nWhen you run this program, it will print the text \"Hello, World!\" to the console.",
  "done": true,
  "done_reason": "stop",
  "total_duration": 4038487077,
  "load_duration": 2965590878,
  "prompt_eval_count": 14,
  "prompt_eval_cached_count": 0,
  "prompt_eval_duration": 54598000,
  "eval_count": 54,
  "eval_duration": 1016789000,
  "tokens_per_second": 53.11,
  "time_to_first_token_ms": 54.6
}

-------------------------------------------------------------------
Process information:
NAME             ID              SIZE     PROCESSOR    CONTEXT    RUNNER      UNTIL              
codestral:22b    0898a8b286d5    14 GB    100% GPU     8192       llamacpp    4 minutes from now    

-------------------------------------------------------------------
GPU memory usage:
Memory used: 14GiB

===================================================================
Running model at 16384 context length...
===================================================================

{
  "model": "codestral:22b",
  "created_at": "2026-10-15T16:42:30.999577057Z",
  "response": " Here is a simple Python program that prints the string \"Hello, World!\" to the console:\n\n```python\nprint(\"Hello, World!\")\n```\n\nWhen you run this program, it will display the text \"Hello, World!\" in the terminal or console.",
  "done": true,
  "done_reason": "stop",
  "total_duration": 4070910346,
  "load_duration": 2893645221,
  "prompt_eval_count": 14,
  "prompt_eval_cached_count": 0,
  "prompt_eval_duration": 54392000,
  "eval_count": 60,
  "eval_duration": 1121394000,
  "tokens_per_second": 53.5,
  "time_to_first_token_ms": 54.39
}

-------------------------------------------------------------------
Process information:
NAME             ID              SIZE     PROCESSOR    CONTEXT    RUNNER      UNTIL              
codestral:22b    0898a8b286d5    16 GB    100% GPU     16384      llamacpp    4 minutes from now    

-------------------------------------------------------------------
GPU memory usage:
Memory used: 16GiB

===================================================================
Running model at 24576 context length...
===================================================================

{
  "model": "codestral:22b",
  "created_at": "2026-10-15T16:42:35.582847329Z",
  "response": " Sure, here's a simple \"Hello, World!\" program written in Python:\n\n```python\nprint(\"Hello, World!\")\n```\n\nThis program will print the text \"Hello, World!\" to the console. You can run this program by saving it to a .py file and executing it using a Python interpreter.",
  "done": true,
  "done_reason": "stop",
  "total_duration": 4319252457,
  "load_duration": 2878782497,
  "prompt_eval_count": 14,
  "prompt_eval_cached_count": 0,
  "prompt_eval_duration": 54618000,
  "eval_count": 74,
  "eval_duration": 1384358000,
  "tokens_per_second": 53.45,
  "time_to_first_token_ms": 54.62
}

-------------------------------------------------------------------
Process information:
NAME             ID              SIZE     PROCESSOR    CONTEXT    RUNNER      UNTIL              
codestral:22b    0898a8b286d5    18 GB    100% GPU     24576      llamacpp    4 minutes from now    

-------------------------------------------------------------------
GPU memory usage:
Memory used: 18GiB

===================================================================
Running model at 32768 context length...
===================================================================

{
  "model": "codestral:22b",
  "created_at": "2026-10-15T16:42:40.006500352Z",
  "response": " Sure, I'd be happy to help with that. Here's a simple \"Hello, World!\" program in Python:\n\n```python\nprint(\"Hello, World!\")\n```\n\nWhen you run this program, it will print the text \"Hello, World!\" to the console.",
  "done": true,
  "done_reason": "stop",
  "total_duration": 4162633943,
  "load_duration": 2899527937,
  "prompt_eval_count": 14,
  "prompt_eval_cached_count": 0,
  "prompt_eval_duration": 54514000,
  "eval_count": 65,
  "eval_duration": 1207123000,
  "tokens_per_second": 53.85,
  "time_to_first_token_ms": 54.51
}

-------------------------------------------------------------------
Process information:
NAME             ID              SIZE     PROCESSOR    CONTEXT    RUNNER      UNTIL              
codestral:22b    0898a8b286d5    20 GB    100% GPU     32768      llamacpp    4 minutes from now    

-------------------------------------------------------------------
GPU memory usage:
Memory used: 19GiB