Compare AI models

Specifications
Parameters
1.05T2.4T
Context window
1M1M
API pricing
Input price
$0.70$2.00International (Singapore)
Output price
$2.10$6.00
Cached input price
$0.07—
Cheapest input
$0.70Mistral$1.15Alibaba
Cheapest output
$2.10Mistral$3.45Alibaba
$245Mistral list price$403Alibaba
Benchmarks
DeepSWE 1.1
61.7%69.3%
Terminal-Bench 4.0
41%26.97%
AutomationBench
59.9%50.8%
Overview
CompanyMistralQwen
Release dateOct 6 2026Sep 2 2026
AccessOpen WeightClosed
Model detailsView modelView model

Frequently asked questions

Mistral Large 4 leads Qwen3.8-Max-0902 on 2 of the 3 benchmarks they both report (DeepSWE 1.1, Terminal-Bench 4.0, AutomationBench). Mistral Large 4 is cheaper on both input and output: $0.70 vs $2.00 per million input tokens, and $2.10 vs $6.00 per million output tokens. The listed Qwen3.8-Max-0902 rates use the “International (Singapore)” pricing scope. Qwen3.8-Max-0902 shipped 34 days before Mistral Large 4, so benchmark comparisons should account for the intervening progress.

Mistral Large 4 has 1.05T parameters, while Qwen3.8-Max-0902 has 2.4T. Context windows are 1M (Mistral Large 4) vs 1M (Qwen3.8-Max-0902). Mistral Large 4 is open weight, while Qwen3.8-Max-0902 is closed.

On DeepSWE 1.1, Qwen3.8-Max-0902 leads at 69.3% vs Mistral Large 4 at 61.7%. On Terminal-Bench 4.0, Mistral Large 4 leads at 41% vs Qwen3.8-Max-0902 at 26.97%. On AutomationBench, Mistral Large 4 leads at 59.9% vs Qwen3.8-Max-0902 at 50.8%.