Compare
Gemini 2.5 Ultra vs Claude Opus 4.7
Google DeepMind vs Anthropic. Specs, benchmarks, and real per-task cost — all in one page.
Verdict
Claude Opus 4.7 leads on LMSYS ELO (1412 vs 1385). Gemini 2.5 Ultra is ~2.9× cheaper on a 3:1 input:output blend.
Google DeepMind
Gemini 2.5 Ultra2M-token context, native video understanding, and Google’s deepest multimodal stack. The long-context king.
2M context windowNative videoBest multimodal reasoning
Anthropic
Claude Opus 4.7Anthropic’s frontier reasoning model. Leads SWE-Bench and holds top ranks on long-context coding and agentic workflows.
Best-in-class coding agent behaviorStrong instruction following1M-token context with high recallHigh reliability under tool use
Pricing
Input / 1M$7.00$15.00
Output / 1M$21.00$75.00
Context2.0M1.0M
Max output64K64K
LicenseProprietaryProprietary
Released2026-02-052026-03-12
Benchmarks
LMSYS ELO1385.01412.0
MMLU Pro90.492.8
HumanEval90.294.4
SWE-Bench58.372.1
MATH92.089.3
GPQA—66.4
MMMU82.179.2
IFEval89.792.1
Per-task cost
Summarize a 1-hour meeting transcript$0.115$0.263
Review a 500-line pull request$0.098$0.270
Answer a customer support ticket$0.038$0.105
Extract structured data from a resume$0.056$0.150
Debug a stack trace with context$0.084$0.240
Per-call cost using published token counts for each task. Real-world prompts vary.