Compare

Gemini 2.5 Ultra vs Claude Sonnet 4.6

Google DeepMind vs Anthropic. Specs, benchmarks, and real per-task cost — all in one page.

Verdict

Gemini 2.5 Ultra leads on LMSYS ELO (1385 vs 1368). Claude Sonnet 4.6 is ~1.8× cheaper on a 3:1 input:output blend.

Google DeepMind

2M-token context, native video understanding, and Google’s deepest multimodal stack. The long-context king.

2M context windowNative videoBest multimodal reasoning

Anthropic

The workhorse. Near-Opus quality at 1/5 the cost. The default choice for production code and agent workloads.

Excellent cost/quality ratioFast inferenceGreat for Cursor/Claude Code

Pricing

Input / 1M$7.00$3.00

Output / 1M$21.00$15.00

Context2.0M500K

Max output64K64K

LicenseProprietaryProprietary

Released2026-02-052025-11-04

Benchmarks

LMSYS ELO1385.01368.0

MMLU Pro90.488.7

HumanEval90.292.1

SWE-Bench58.364.5

MATH92.085.0

GPQA—59.4

MMMU82.174.1

IFEval89.790.5

Per-task cost

Summarize a 1-hour meeting transcript$0.115$0.052

Review a 500-line pull request$0.098$0.054

Answer a customer support ticket$0.038$0.021

Extract structured data from a resume$0.056$0.030

Debug a stack trace with context$0.084$0.048

Per-call cost using published token counts for each task. Real-world prompts vary.

Add up to four models, tweak tasks live.