SShogo Bench

Mistral Large 4

Mistral's 1T-parameter multimodal MoE for reasoning, coding, and agentic work

Shogo Agent Index

Unranked

Not ranked yet: 1 agentic benchmark measured, and we need at least 2 plus an overall capability score. It will be ranked as results are published.

Agentic coding
48
estimated
Long-horizon autonomy
53
estimated
Knowledge-work agents
39
32% measured
Computer use
32
estimated
Security & context learning
45
estimated

Price

Input
$0.70/M
Output
$2.10/M
Cache read
$0.070/M
3:1 blend
$1.05/M
Agent-effective
$0.26/M

Cheapest host for agents: Mistral at $0.26/M.

Agent-effective = 20 input tokens per output token, 85% cache hits. Why.

Agentic benchmarks

BenchmarkScoreNormalisedIndex weightHarness / source
SWE-bench Verifiedest.8%predicted 75
Terminal-Bench 2.0est.8%predicted 64
DeepSWEest.6%predicted 43
FrontierSWEest.5%predicted 12
CursorBenchest.4%predicted 35
GSO-Benchest.4%predicted 23
METR Time Horizonsest.10%predicted 73
Vending-Bench 2est.6%predicted 38
PostTrainBenchest.4%predicted 23
APEX-Agents51.6%8%source
GDPvalest.7%predicted 43
Remote Labor Indexest.5%predicted 3
DeepResearch Benchest.5%predicted 47
OSWorld 2.0est.6%predicted 7
OSWorldest.4%predicted 70
Cybenchest.5%predicted 69
CL-benchest.5%predicted 21

Other benchmarks

BenchmarkCategoryScoreSource
Chess PuzzlesReasoning33.0%source
CritPtReasoning11.1%source
DTBenchKnowledge88.2%source
FrontierCodeCoding39.4%chisel
FrontierMath Tiers 1-3Math58.6%source
FrontierMath Tier 4Math26.8%source
GPQA DiamondReasoning90.2%source
LMCAKnowledge35.8%source
MirrorCodeCoding40.0%source
OTIS Mock AIMEMath99.4%source
ProofBenchMath10.0%source
SciCodeCoding54.2%source
SimpleQA VerifiedKnowledge22.4%source
WeirdML v3Coding4.2%opencode

API hosts

Live prices and uptime for every host serving Mistral Large 4 via OpenRouter.

HostInputOutputCache readAgent $/MUptime 24hQuant
Mistral$0.70$2.10$0.070$0.2699.7%—
Mistral$0.77$2.31$0.077$0.2899.0%—