GPT-4o temperature 0 still gives different outputs

I set temperature: 0 for reproducible outputs but I'm still getting slightly different responses for the exact same prompt. Is this expected?

for i in range(5):
    response = client.chat.completions.create(
        model="gpt-4o",
        messages=[{"role": "user", "content": "Write exactly 3 bullet points about Python"}],
        temperature=0,
        seed=42
    )
    print(response.choices[0].message.content)
    print(response.system_fingerprint)
    print("---")

Even with the same seed, outputs vary. The system_fingerprint also changes between calls. Is temperature 0 not truly deterministic?