Compare AI models

Specifications
Parameters
1.05T744B
Context window
1M200k
API pricing
Input price
$0.70$1.00
Output price
$2.10$3.20
Cached input price
$0.07$0.20
Cheapest input
$0.70Mistral$0.60StreamLake
Cheapest output
$2.10Mistral$1.92StreamLake
$245Mistral list price$216StreamLake · fp8
Benchmarks
BullshitBench v2
80%28%
BrowseComp
84.5%62%
Overview
CompanyMistralZ.ai
Release dateOct 6 2026Feb 11 2026
AccessOpen WeightOpen Weight
Model detailsView modelView model

Frequently asked questions

Mistral Large 4 leads GLM-5 on 2 of the 2 benchmarks they both report (BullshitBench v2, BrowseComp). Mistral Large 4 is cheaper on both input and output: $0.70 vs $1.00 per million input tokens, and $2.10 vs $3.20 per million output tokens. GLM-5 shipped 237 days before Mistral Large 4, so benchmark comparisons should account for the intervening progress.

Mistral Large 4 has 1.05T parameters, while GLM-5 has 744B. Context windows are 1M (Mistral Large 4) vs 200k (GLM-5).

On BullshitBench v2, Mistral Large 4 leads at 80% vs GLM-5 at 28%. On BrowseComp, Mistral Large 4 leads at 84.5% vs GLM-5 at 62%.