Compare AI models

Specifications
Parameters
1.05T320B
Context window
1M1M
API pricing
Input price
$0.70—
Output price
$2.10—
Cached input price
$0.07—
Cheapest input
$0.70Mistral$0.0399Relace
Cheapest output
$2.10Mistral$0.0784Relace
$245Mistral list price$11.90Relace · fp4
Benchmarks
DeepSWE 1.1
61.7%63.4%
Terminal-Bench 4.0
41%35.76%
AutomationBench
59.9%48.8%
threejseval
10081378
Overview
CompanyMistralZ.ai
Release dateOct 6 2026Aug 26 2026
AccessOpen WeightOpen Weight
Model detailsView modelView model

Frequently asked questions

Mistral Large 4 and GLM-5.3-Flash are evenly matched across the 4 benchmarks they both report (DeepSWE 1.1, Terminal-Bench 4.0, AutomationBench, threejseval). 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 GLM-5.3-Flash. GLM-5.3-Flash shipped 41 days before Mistral Large 4, so benchmark comparisons should account for the intervening progress.

Mistral Large 4 has 1.05T parameters, while GLM-5.3-Flash has 320B. Context windows are 1M (Mistral Large 4) vs 1M (GLM-5.3-Flash).

On DeepSWE 1.1, GLM-5.3-Flash leads at 63.4% vs Mistral Large 4 at 61.7%. On Terminal-Bench 4.0, Mistral Large 4 leads at 41% vs GLM-5.3-Flash at 35.76%. On AutomationBench, Mistral Large 4 leads at 59.9% vs GLM-5.3-Flash at 48.8%. On threejseval, GLM-5.3-Flash leads at 1378 vs Mistral Large 4 at 1008.