Compare

Claude Sonnet 4.6 vs Gemini 2.5 Flash

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

Verdict

Claude Sonnet 4.6 leads on LMSYS ELO (1368 vs 1312). Gemini 2.5 Flash is ~11.4× cheaper on a 3:1 input:output blend.

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

Google DeepMind

Google’s price/performance darling. 1M context at $0.30/M input — nobody comes close on throughput-per-dollar.

Absurd cost-to-context ratioFast multimodalFree tier via AI Studio

Pricing

Input / 1M$3.00$0.30

Output / 1M$15.00$1.20

Context500K1.0M

Max output64K32K

LicenseProprietaryProprietary

Released2025-11-042026-02-05

Benchmarks

LMSYS ELO1368.01312.0

MMLU Pro88.782.1

HumanEval92.185.8

SWE-Bench64.5—

MATH85.084.2

GPQA59.4—

MMMU74.174.6

IFEval90.586.1

Per-task cost

Summarize a 1-hour meeting transcript$0.052$0.0051

Review a 500-line pull request$0.054$0.0048

Answer a customer support ticket$0.021$0.0019

Extract structured data from a resume$0.030$0.0027

Debug a stack trace with context$0.048$0.0042

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

Add up to four models, tweak tasks live.