Compare AI models

Specifications
Parameters
671B4B/30B/235B
Context window
128k256k
API pricing
Cheapest input
$0.50SiliconFlow—
Cheapest output
$2.18SiliconFlow—
$209SiliconFlow · fp8—
Benchmarks
LiveCodeBench
73.3%74.1%
Humanity's Last Exam · text only, no tools
17.7%18.2%
AIME 2025
87.5%92.3%
GPQA Diamond
81%81.1%
MMLU-Pro
85%84.4%
Overview
CompanyDeepSeekQwen
Release dateMay 28 2025Jul 21 2025
AccessOpen WeightOpen Weight
Model detailsView modelView model

Frequently asked questions

Qwen3-2507 leads DeepSeek R1-0528 on 4 of the 5 benchmarks they both report. DeepSeek R1-0528 shipped 54 days before Qwen3-2507, so benchmark comparisons should account for the intervening progress.

DeepSeek R1-0528 has 671B parameters, while Qwen3-2507 has 4B/30B/235B. Context windows are 128k (DeepSeek R1-0528) vs 256k (Qwen3-2507).

On LiveCodeBench, Qwen3-2507 leads at 74.1% vs DeepSeek R1-0528 at 73.3%. On Humanity's Last Exam · text only, no tools, Qwen3-2507 leads at 18.2% vs DeepSeek R1-0528 at 17.7%. On AIME 2025, Qwen3-2507 leads at 92.3% vs DeepSeek R1-0528 at 87.5%. On GPQA Diamond, Qwen3-2507 leads at 81.1% vs DeepSeek R1-0528 at 81%. On MMLU-Pro, DeepSeek R1-0528 leads at 85% vs Qwen3-2507 at 84.4%.