AiCompareNet
Zurück zur Entwicklerzone

Mistral Modelle

Entdecken Sie alle 3 Modelle von Mistral mit detaillierten Preisen, Vor- und Nachteilen sowie Entwicklerempfehlungen.

3
Modelle
$0.400
Niedrigster Input
131K
Max. Kontext
2
Qualitätsstufen

Schnellempfehlungen

Bestes Preis-Leistungs-Verhältnis: Mistral Medium 3 ($0.400/1M)
Beste Qualität: Mistral Large 2

Mistral Large 2

Flagship

Multilingual, complex tasks

Offizielle Preise

Wann verwenden: Top pick for multilingual enterprise apps, especially EU-based deployments.

Upgrade-Highlights

  • ◆MMLU: 84.0% — competitive with GPT-4 class models
  • ◆Fine-tuning available — customize for domain-specific tasks
  • ◆EU-based (GDPR compliant) — only major flagship with EU sovereignty
  • ◆12 languages optimized — best-in-class for French, German, Spanish
  • ◆128K context is smaller vs 1M competitors — trade-off for EU compliance
Input-Preis
$2.00
per 1M tokens
Output-Preis
$6.00
per 1M tokens
Cached Input
—
per 1M tokens
Batch-Input
—
per 1M tokens
Kontextfenster: 131K
Max. Output: 8,192 tokens
Wissensstand: 2024-07
VisionFunktionsaufrufFeinabstimmungJSON-ModusKostenlose Stufe

Vorteile

  • Excellent multilingual (French, German, etc.)
  • Fine-tuning available
  • EU-based (GDPR friendly)

Nachteile

  • No vision
  • 128K context is small vs competitors
  • No cached/batch pricing

Leistung

Ausgabegeschwindigkeit~55 tok/s
Rate-Limit3,000 RPM

Multimodal

BildeingabeBildausgabeAudioeingabeAudioausgabe

Benchmarks

MMLU
84.0%
HumanEval
82.0%
MATH
65.0%

Mistral Medium 3

Mid-tier

Balanced multilingual

Offizielle Preise

Wann verwenden: Cost-effective multilingual + vision for European market applications.

Upgrade-Highlights

  • ◆Vision added at mid-tier — multimodal capability at $0.40/M input
  • ◆5x cheaper than Mistral Large ($0.40 vs $2/M) for lighter tasks
  • ◆EU-based infrastructure — GDPR compliant for European data
  • ◆128K context — sufficient for most enterprise document processing
  • ◆No fine-tuning — upgrade to Large 2 for custom model training
Input-Preis
$0.400
per 1M tokens
Output-Preis
$2.00
per 1M tokens
Cached Input
—
per 1M tokens
Batch-Input
—
per 1M tokens
Kontextfenster: 131K
Max. Output: 8,192 tokens
Wissensstand: 2024-07
VisionFunktionsaufrufFeinabstimmungJSON-ModusKostenlose Stufe

Vorteile

  • Vision at mid-tier price
  • Good multilingual balance
  • EU-based

Nachteile

  • No fine-tuning
  • No cached/batch pricing
  • Smaller context than competitors

Leistung

Ausgabegeschwindigkeit~75 tok/s
Rate-Limit5,000 RPM

Multimodal

BildeingabeBildausgabeAudioeingabeAudioausgabe

Benchmarks

MMLU
80.5%
HumanEval
78.0%

Mixtral 8x22B

Mid-tier

Open-weight, high throughput

Offizielle Preise

Wann verwenden: For self-hosting or fine-tuning on domain-specific data with high throughput needs.

Upgrade-Highlights

  • ◆Open-weight MoE: 8x22B params, only 39B active per token — high throughput
  • ◆Fine-tunable — full weight access for domain adaptation
  • ◆Function calling + JSON mode — enterprise-ready tool integration
  • ◆65K context — smaller than newer models but sufficient for most tasks
  • ◆Self-hostable — no per-token cost when running on own infrastructure
Input-Preis
$0.900
per 1M tokens
Output-Preis
$2.70
per 1M tokens
Cached Input
—
per 1M tokens
Batch-Input
—
per 1M tokens
Kontextfenster: 66K
Max. Output: 4,096 tokens
Wissensstand: 2024-01
VisionFunktionsaufrufFeinabstimmungJSON-ModusKostenlose Stufe

Vorteile

  • Open-weight MoE architecture
  • Fine-tunable
  • High throughput via sparse activation

Nachteile

  • Only 65K context
  • No vision
  • Older knowledge cutoff (2024-01)

Leistung

Ausgabegeschwindigkeit~85 tok/s
Rate-Limit—

Multimodal

BildeingabeBildausgabeAudioeingabeAudioausgabe

Benchmarks

MMLU
77.8%
HumanEval
75.5%
MATH
58.0%

Nebeneinander-Vergleich

ModellStufeInputOutputKontext
Mistral Large 2Flagship$2.00$6.00131K
Mistral Medium 3Mid-tier$0.400$2.00131K
Mixtral 8x22BMid-tier$0.900$2.7066K