Compare AI models

Specifications
Parameters
1.05T—
Context window
1M1.05M
API pricing
Input price
$0.70$0.10
Output price
$2.10$0.50
Cached input price
$0.07$0.01
Cheapest input
$0.70Mistral$0.05OpenAI
Cheapest output
$2.10Mistral$0.25OpenAI
$245Mistral list price$45.00OpenAI list price
Benchmarks
BullshitBench v2
80%59%
DeepSWE 1.1
61.7%66.6%
Terminal-Bench 4.0
41%16.36%
threejseval
10081365
Overview
CompanyMistralOpenAI
Release dateOct 6 2026Sep 22 2026
AccessOpen WeightClosed
Model detailsView modelView model

Frequently asked questions

Mistral Large 4 and GPT-6 Luna are evenly matched across the 4 benchmarks they both report (BullshitBench v2, DeepSWE 1.1, Terminal-Bench 4.0, threejseval). GPT-6 Luna is cheaper on both input and output: $0.10 vs $0.70 per million input tokens, and $0.50 vs $2.10 per million output tokens. Figures are base-tier rates. GPT-6 Luna shipped 14 days before Mistral Large 4, so benchmark comparisons should account for the intervening progress.

Context windows are 1M (Mistral Large 4) vs 1.05M (GPT-6 Luna). Mistral Large 4 is open weight, while GPT-6 Luna is closed.

On BullshitBench v2, Mistral Large 4 leads at 80% vs GPT-6 Luna at 59%. On DeepSWE 1.1, GPT-6 Luna leads at 66.6% vs Mistral Large 4 at 61.7%. On Terminal-Bench 4.0, Mistral Large 4 leads at 41% vs GPT-6 Luna at 16.36%. On threejseval, GPT-6 Luna leads at 1365 vs Mistral Large 4 at 1008.