Compare AI models

Specifications
Parameters
—1.05T
Context window
1M1M
API pricing
Input price
$3.00$0.70
Output price
$15.00$2.10
Cached input price
$0.30$0.07
Cheapest input
$3.00Amazon Bedrock$0.70Mistral
Cheapest output
$15.00Amazon Bedrock$2.10Mistral
$1,350Anthropic list price$245Mistral list price
Benchmarks
BullshitBench v2
91%80%
DeepSWE 1.1
30%61.7%
BrowseComp
76.2%84.5%
AutomationBench
5.3%59.9%
Overview
CompanyAnthropicMistral
Release dateFeb 17 2026Oct 6 2026
AccessClosedOpen Weight
Model detailsView modelView model

Frequently asked questions

Mistral Large 4 leads Claude Sonnet 4.6 on 3 of the 4 benchmarks they both report (BullshitBench v2, DeepSWE 1.1, BrowseComp, AutomationBench). Mistral Large 4 is cheaper on both input and output: $0.70 vs $3.00 per million input tokens, and $2.10 vs $15.00 per million output tokens. Claude Sonnet 4.6 shipped 231 days before Mistral Large 4, so benchmark comparisons should account for the intervening progress.

Context windows are 1M (Claude Sonnet 4.6) vs 1M (Mistral Large 4). Claude Sonnet 4.6 is closed, while Mistral Large 4 is open weight.

On BullshitBench v2, Claude Sonnet 4.6 leads at 91% vs Mistral Large 4 at 80%. On DeepSWE 1.1, Mistral Large 4 leads at 61.7% vs Claude Sonnet 4.6 at 30%. On BrowseComp, Mistral Large 4 leads at 84.5% vs Claude Sonnet 4.6 at 76.2%. On AutomationBench, Mistral Large 4 leads at 59.9% vs Claude Sonnet 4.6 at 5.3%.