Update the models metadata

[ArtyomVancyan]

This PR updates the models metadata to keep the repository up-to-date with the upstream model costs.

[ArtyomVancyan]

@copilot, please fix the following issue and update this PR:

tests/lib/test_formatter.py ................................F..F.        [ 78%]
tests/lib/test_topology.py .....                                         [ 89%]
tests/lib/test_watching.py .....                                         [100%]

=================================== FAILURES ===================================
____________________ TestCosts.test_known_model_with_tokens ____________________

self = <lib.test_formatter.TestCosts object at 0x7f30b9925b10>

    def test_known_model_with_tokens(self):
        result = Formatter.costs("openai/gpt-4o-mini", 1000, 2000)
>       assert result == {"cached": "0.0", "total": "0.0"}
E       AssertionError: assert {'cached': '0...al': '0.0012'} == {'cached': '0...total': '0.0'}
E         
E         Differing items:
E         {'total': '0.0012'} != {'total': '0.0'}
E         {'cached': '0.00008'} != {'cached': '0.0'}
E         Use -v to get more diff

tests/lib/test_formatter.py:307: AssertionError
____________________ TestMetrics.test_cost_for_known_model _____________________

self = <lib.test_formatter.TestMetrics object at 0x7f30b9924040>

    def test_cost_for_known_model(self):
        run = make_metrics_run(
            extra={"ls_model_name": "openai/gpt-4o-mini"},
            outputs={"total_tokens": 2000, "cached_tokens": 1000},
        )
        result = Formatter.metrics(run)
>       assert result["costs"] == {"cached": "0.0", "total": "0.0"}
E       AssertionError: assert {'cached': '0...al': '0.0012'} == {'cached': '0...total': '0.0'}
E         
E         Differing items:
E         {'total': '0.0012'} != {'total': '0.0'}
E         {'cached': '0.00008'} != {'cached': '0.0'}
E         Use -v to get more diff

tests/lib/test_formatter.py:327: AssertionError
=========================== short test summary info ============================
FAILED tests/lib/test_formatter.py::TestCosts::test_known_model_with_tokens
FAILED tests/lib/test_formatter.py::TestMetrics::test_cost_for_known_model - ...
========================= 2 failed, 45 passed in 2.83s =========================

[Copilot]

@copilot, please fix the following issue and update this PR:

...

Fixed in 59bb0d7. The PR updated openai/gpt-4o-mini costs from zero to cache_read: 0.08, output: 0.6, so I updated both test assertions to reflect the actual computed costs: {"cached": "0.00008", "total": "0.0012"}.