Compare AI models

Specifications
Parameters
284B1.05T
Context window
1M1M
API pricing
Input price
—$0.70
Output price
—$2.10
Cached input price
—$0.07
Cheapest input
$0.044Baidu$0.70Mistral
Cheapest output
$0.132Baidu$2.10Mistral
$15.40Baidu · fp8$245Mistral list price
Benchmarks
BullshitBench v2
39%80%
BrowseComp
73.2%84.5%
AutomationBench
25.1%59.9%
Overview
CompanyDeepSeekMistral
Release dateJul 31 2026Oct 6 2026
AccessOpen WeightOpen Weight
Model detailsView modelView model

Frequently asked questions

Mistral Large 4 leads DeepSeek-V4-Flash-0731 on 3 of the 3 benchmarks they both report (BullshitBench v2, BrowseComp, AutomationBench). Only Mistral Large 4 has a verified first-party price for an available API: $0.70 per million input tokens and $2.10 per million output tokens. No currently available pay-as-you-go API rate is tracked for DeepSeek-V4-Flash-0731. DeepSeek-V4-Flash-0731 shipped 67 days before Mistral Large 4, so benchmark comparisons should account for the intervening progress.

DeepSeek-V4-Flash-0731 has 284B parameters, while Mistral Large 4 has 1.05T. Context windows are 1M (DeepSeek-V4-Flash-0731) vs 1M (Mistral Large 4).

On BullshitBench v2, Mistral Large 4 leads at 80% vs DeepSeek-V4-Flash-0731 at 39%. On BrowseComp, Mistral Large 4 leads at 84.5% vs DeepSeek-V4-Flash-0731 at 73.2%. On AutomationBench, Mistral Large 4 leads at 59.9% vs DeepSeek-V4-Flash-0731 at 25.1%.