invokeable()
← comparisons / claude-opus-5-vs-gemini-3.7-flash

Claude Opus 5 vs Gemini 3.7 Flash

Side-by-side from the Invokeable directory — provider-attributed pricing, canonical IDs, draft data flagged. Open as a shareable card →

01Head to head

Claude Opus 5Gemini 3.7 Flash
CreatorAnthropicGoogle
Released2026-07-242026-08-13
Context1M1M
Max output128K66K
Input $/M——
Output $/M——
Knowledge cutoff2026-052026-03
Modalitiestext + image → texttext + image + video + audio → text
Capabilitiestool-use, extended-thinkingtool-use, extended-thinking, structured-output
ProvidersAnthropic APIGemini API (AI Studio)

02Benchmarks · vendor-reported unless verified

Claude Opus 5Gemini 3.7 Flash
AutomationBench (accuracy)—30.4
DeepSWE68.8%65.3%
FrontierCode (mean@5)53.4—
FrontierCode (score)—43.6
GDM-MRCR (accuracy)—97
GDP.pdf (accuracy)—34
SWE-bench Multilingual89.5%—
SWE-bench Multimodal59.4%—
SWE-bench Pro79.2%—
SWE-bench Verified (resolved)96—
Terminal-Bench (accuracy)—85.8