Compare

Gemini 2.5 Flash vs DeepSeek V4

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

Verdict

DeepSeek V4 leads on LMSYS ELO (1334 vs 1312). DeepSeek V4 is ~3.0× cheaper on a 3:1 input:output blend. DeepSeek V4 is open-weights, the other is proprietary.

Google DeepMind

Gemini 2.5 Flash

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

DeepSeek

DeepSeek V4

The shock-the-market moment of 2026. Frontier coding quality at 1/50th the price of Opus.

Unbeatable $/qualityOpen weightsGreat at code

Pricing

Input / 1M$0.30$0.14

Output / 1M$1.20$0.28

Context1.0M128K

Max output32K16K

LicenseProprietaryOpen weights

Released2026-02-052026-02-20

Benchmarks

LMSYS ELO1312.01334.0

MMLU Pro82.185.4

HumanEval85.891.5

SWE-Bench—54.1

MATH84.288.7

MMMU74.6—

IFEval86.185.8

Per-task cost

Summarize a 1-hour meeting transcript$0.0051$0.0022

Review a 500-line pull request$0.0048$0.0017

Answer a customer support ticket$0.0019$0.0006

Extract structured data from a resume$0.0027$0.0010

Debug a stack trace with context$0.0042$0.0014

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

Tool

Interactive compare →

Add up to four models, tweak tasks live.