Compare models

Claude Sonnet 5 vs Gemini 3.7 Flash 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.
Claude Sonnet 5 vs Gemini 3.7 Flash vs Phi-4, compared across 14 tracked fields. Each value shows the source that reported it and when.
FieldClaude Sonnet 5Gemini 3.7 FlashPhi-4
Downloads, 30 days——
591,256Synced from sourceCurrent4h ago
Likes——
2,312Synced from sourceCurrent4h ago
Context window
1,000,000Synced from sourceCurrent4h ago
1,048,576 (highest of the compared values)Synced from sourceCurrent4h ago
16,384Synced from sourceCurrent4h ago
Input, $/M tokens
2Synced from sourceCurrent4h ago
0.75Synced from sourceCurrent4h ago
0.07 (lowest of the compared values)Synced from sourceCurrent4h ago
Modality
textSeeded, unreviewedwritten 1mo ago
textSeeded, unreviewedwritten 1mo ago
textSeeded, unreviewedwritten 1mo ago
Licence
proprietarySeeded, unreviewedwritten 1mo ago
proprietarySeeded, unreviewedwritten 1mo ago
mitSeeded, unreviewedwritten 1mo ago
Gated——
noSynced from sourceCurrent4h ago
Last updated——
2026-07-14Synced from sourceCurrent4h ago
Library——
transformersSynced from sourceCurrent4h ago
Task——
text-generationSynced from sourceCurrent4h ago
Hugging Face——
microsoft/phi-4Synced from sourceCurrent4h ago
OpenRouter id
anthropic/claude-sonnet-5Synced from sourceCurrent4h ago
google/gemini-3.7-flashSynced from sourceCurrent4h ago
microsoft/phi-4Synced from sourceCurrent4h ago
Output, $/M tokens
10Synced from sourceCurrent4h ago
3.75Synced from sourceCurrent4h ago
0.14Synced from sourceCurrent4h ago
Vendor
AnthropicSeeded, unreviewedwritten 1mo ago
GoogleSeeded, unreviewedwritten 1mo ago
MicrosoftSeeded, unreviewedwritten 1mo ago

Sources: Hugging Face Hub, OpenRouter model catalogue. Most recent observation 4h 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.