Mistral AIopen-weight
Released: 2026-06-01

Mistral Large

This is Mistral AI's flagship model, Mistral Large 2 (version `mistral-large-2407`). It's a proprietary weights-available model and excels at reasoning, code, JSON, chat, and more. Read the launch announcement [here](https://mistral.ai/news/mistral-large-2407/)....

Intelligence Index
88.5/ 100
Calibrated multi-domain composite
Throughput Speed
55tok/sec
Output streaming throughput
Time To First Token
420ms
Initial chunk server latency
LiveBench Contam-Free
77.9%
September 2026 suite score

LiveBench Multi-Domain Evaluation Breakdown

Monthly refreshed contamination-free test sets
Logical Reasoning81.4%
Coding & Repo Repair79.7%
Mathematics (AIME)78.8%
Data Analysis & Tables75.2%
Language & Comprehension77.0%
Instruction Following77.9%

Sourced Benchmark Evaluations (61)

Verified performance across authoritative benchmarks with provenance tracking.
BenchmarkStatusScoreTrustDateSource TypeProvenance
ARC-AGI-3Active55.8%
93
2026-07-06independentSource ↗
ARC-AGI-2Nearing Saturation66.9%
75
2026-06-29independentSource ↗
FrontierMathActive23%
98
2026-06-26independentSource ↗
PutnamBenchActive42.8%
90
2026-06-25independentSource ↗
SWE-LancerDeprecated57.2%
20
2026-06-23independentSource ↗
Omni-MATHDeprecated60.9%
10
2026-06-23independentSource ↗
SoohakActive62.5%
98
2026-06-21independentSource ↗
FrontierCodeActive48.8%
98
2026-06-21independentSource ↗
Humanity's Last ExamActive37.8%
93
2026-06-19independentSource ↗
StructFlowBenchNearing Saturation66.2%
75
2026-06-19independentSource ↗
AIMENearing Saturation78.2%
75
2026-06-19independentSource ↗
OlympiadBenchActive67.5%
80
2026-06-18independentSource ↗
MathArenaActive65.8%
99
2026-06-17independentSource ↗
FireBenchActive65.4%
95
2026-06-17independentSource ↗
Terminal-BenchActive58.9%
80
2026-06-17independentSource ↗
OpenAI-MRCRActive74.3%
80
2026-06-16independentSource ↗
SimpleQAActive46%
90
2026-06-16independentSource ↗
CodeContestsDeprecated54.7%
10
2026-06-16independentSource ↗
ComplexBenchNearing Saturation63.9%
80
2026-06-15independentSource ↗
SWE-bench ProActive52.1%
90
2026-06-15independentSource ↗
LongBench-ProActive72.6%
80
2026-06-15independentSource ↗
ZeroSCROLLSNearing Saturation70.8%
78
2026-06-14independentSource ↗
MultiPL-EDeprecated77.4%
10
2026-06-14independentSource ↗
LiveCodeBenchSaturated69%
39
2026-06-13independentSource ↗
AgentIFNearing Saturation53%
80
2026-06-13independentSource ↗
MATHSaturated80.7%
25
2026-06-13independentSource ↗
InfiniteBenchNearing Saturation75.2%
65
2026-06-13independentSource ↗
MercuryDeprecated74%
10
2026-06-12independentSource ↗
SCROLLSNearing Saturation70.8%
65
2026-06-12independentSource ↗
MMLU-ProNearing Saturation81.4%
75
2026-06-12independentSource ↗
RULERNearing Saturation81.4%
84
2026-06-11independentSource ↗
TabMWPSaturated75.7%
25
2026-06-11independentSource ↗
CyberSecEvalDeprecated75.7%
10
2026-06-11independentSource ↗
SWE-bench VerifiedSaturated65.6%
25
2026-06-10independentSource ↗
MMLU-ReduxSaturated86.7%
25
2026-06-10independentSource ↗
QuALITYNearing Saturation77.9%
78
2026-06-10independentSource ↗
LongBenchNearing Saturation76.1%
65
2026-06-09independentSource ↗
SWE-benchDeprecated57.2%
10
2026-06-09independentSource ↗
MGSMNearing Saturation78.2%
65
2026-06-09independentSource ↗
AGIEvalSaturated73.8%
25
2026-06-08independentSource ↗
QASPERNearing Saturation73.5%
65
2026-06-08independentSource ↗
TruthfulQASaturated77.9%
25
2026-06-08independentSource ↗
GSM-HardDeprecated79%
10
2026-06-07independentSource ↗
SuperGLUESaturated87.5%
25
2026-06-07independentSource ↗
CodeXGLUEDeprecated79.1%
23
2026-06-06independentSource ↗
MT-BenchNearing Saturation8%
75
2026-06-06independentSource ↗
BIG-Bench HardSaturated84.1%
25
2026-06-06independentSource ↗
WinoGrandeSaturated89.2%
28
2026-06-05independentSource ↗
MBPP+Nearing Saturation85.8%
65
2026-06-05independentSource ↗
IFEvalDeprecated76.3%
25
2026-06-05independentSource ↗
GPQANearing Saturation77.9%
78
2026-06-05independentSource ↗
Needle-in-a-HaystackSaturated97.7%
44
2026-06-04independentSource ↗
HumanEval+Saturated90.8%
25
2026-06-04independentSource ↗
CommonsenseQASaturated90.1%
28
2026-06-04independentSource ↗
MBPPSaturated89.1%
25
2026-06-03independentSource ↗
GSM8KSaturated88.9%
25
2026-06-03independentSource ↗
GLUESaturated92.7%
28
2026-06-03independentSource ↗
MMLUSaturated90.3%
25
2026-06-02independentSource ↗
HumanEvalSaturated95.9%
25
2026-06-02independentSource ↗
ARC-AGISaturated80.7%
25
2026-06-02independentSource ↗
LiveBench (Continuous)Active77.9%
94
2025-02-01independentSource ↗
Local Hardware Execution (Ollama)Hardware Compatible
# 1. Pull and execute model locally with Ollama
ollama run mistral-large

# 2. Or invoke via local OpenAI-compatible endpoint
curl http://localhost:11434/v1/chat/completions \
  -H "Content-Type: application/json" \
  -d '{
    "model": "mistral-large",
    "messages": [{"role": "user", "content": "Analyze reasoning chains on AIME 2026"}]
  }'

Evaluation & Sourcing Notes

Scores listed for Mistral Large represent verified evaluations extracted from official research papers, independent evaluation suites (HELM, LMSYS, OpenCompass, LiveBench), and verified audit reports.

All benchmarks marked as saturated or deprecated reflect historical performance where the benchmark no longer provides active discriminative power.