Compare AI models

Specifications
Parameters
—1.05T
Context window
1M1M
API pricing
Input price
$2.00$0.70
Output price
$12.00$2.10
Cached input price
$0.20$0.07
Cheapest input
$1.00Google AI Studio$0.70Mistral
Cheapest output
$6.00Google AI Studio$2.10Mistral
$1,000Google list price$245Mistral list price
Benchmarks
BullshitBench v2
37%80%
DeepSWE 1.1
12%61.7%
Terminal-Bench 4.0
2.73%41%
BrowseComp
85.9%84.5%
Overview
CompanyGoogleMistral
Release dateFeb 19 2026Oct 6 2026
AccessClosedOpen Weight
Model detailsView modelView model

Frequently asked questions

Mistral Large 4 leads Gemini 3.1 Pro on 3 of the 4 benchmarks they both report (BullshitBench v2, DeepSWE 1.1, Terminal-Bench 4.0, BrowseComp). Mistral Large 4 is cheaper on both input and output: $0.70 vs $2.00 per million input tokens, and $2.10 vs $12.00 per million output tokens. Figures are base-tier rates. Gemini 3.1 Pro shipped 229 days before Mistral Large 4, so benchmark comparisons should account for the intervening progress.

Context windows are 1M (Gemini 3.1 Pro) vs 1M (Mistral Large 4). Gemini 3.1 Pro is closed, while Mistral Large 4 is open weight.

On BullshitBench v2, Mistral Large 4 leads at 80% vs Gemini 3.1 Pro at 37%. On DeepSWE 1.1, Mistral Large 4 leads at 61.7% vs Gemini 3.1 Pro at 12%. On Terminal-Bench 4.0, Mistral Large 4 leads at 41% vs Gemini 3.1 Pro at 2.73%. On BrowseComp, Gemini 3.1 Pro leads at 85.9% vs Mistral Large 4 at 84.5%.