Compare models

GPT-4o vs Claude Sonnet 5 vs Phi-4

Every figure below carries the source that reported it and the moment it was observed. Values that failed validation are withheld rather than shown, and a dash means no source we track carries that field for that model.
GPT-4o vs Claude Sonnet 5 vs Phi-4, compared across 14 tracked fields. Each value shows the source that reported it and when.
FieldGPT-4oClaude Sonnet 5Phi-4
Downloads, 30 days
624,150Synced from sourceCurrent3h ago
Likes
2,309Synced from sourceCurrent3h ago
Context window
128,000Synced from sourceCurrent3h ago
1,000,000 (highest of the compared values)Synced from sourceCurrent3h ago
16,384Synced from sourceCurrent3h ago
Input, $/M tokens
2.5Synced from sourceCurrent3h ago
2Synced from sourceCurrent3h ago
0.07 (lowest of the compared values)Synced from sourceCurrent3h ago
Modality
textSeeded, unreviewedwritten 29d ago
textSeeded, unreviewedwritten 29d ago
textSeeded, unreviewedwritten 29d ago
Licence
proprietarySeeded, unreviewedwritten 29d ago
proprietarySeeded, unreviewedwritten 29d ago
mitSeeded, unreviewedwritten 29d ago
Gated
noSynced from sourceCurrent3h ago
Last updated
2026-07-14Synced from sourceCurrent3h ago
Library
transformersSynced from sourceCurrent3h ago
Task
text-generationSynced from sourceCurrent3h ago
Hugging Face
microsoft/phi-4Synced from sourceCurrent3h ago
OpenRouter id
openai/gpt-4oSynced from sourceCurrent3h ago
anthropic/claude-sonnet-5Synced from sourceCurrent3h ago
microsoft/phi-4Synced from sourceCurrent3h ago
Output, $/M tokens
10Synced from sourceCurrent3h ago
10Synced from sourceCurrent3h ago
0.14Synced from sourceCurrent3h ago
Vendor
OpenAISeeded, unreviewedwritten 29d ago
AnthropicSeeded, unreviewedwritten 29d ago
MicrosoftSeeded, unreviewedwritten 29d ago

Sources: Hugging Face Hub, OpenRouter model catalogue. Most recent observation 3h ago. Where a row is marked, ▲ is the highest of the values shown and ▼ the lowest — arithmetic on the figures above, not a ranking or a recommendation. Rows where neither extreme is meaningful are left unmarked rather than given a direction they do not have.

Change the comparison

Remove one, or open the directory to pick a different set. Up to 3 at a time.


Every figure on this site resolves to the source that reported it, the moment it was observed, and how it was measured. A value that fails validation is withheld rather than shown — withheld means we refused to publish it, not that it is zero.