Zum Inhalt springen
Kuratierte KI-Intelligenz Est. 2026
Technical Curator Live

Model Registry

search
leaderboard Arena ELO

15 Modelle

deepseek-chat

DeepSeek: DeepSeek V3

memory 164k
stable $0.320 / $0.890 /1M

deepseek-r1-distill-llama-70b

DeepSeek: R1 Distill Llama 70B

memory 131k
stable $0.700 / $0.800 /1M

deepseek-r1-distill-qwen-32b

DeepSeek: R1 Distill Qwen 32B

memory 33k
stable $0.290 / $0.290 /1M

deepseek-chat-v3-0324

DeepSeek: DeepSeek V3 0324

memory 164k
stable $0.200 / $0.770 /1M

deepseek-chat-v3.1

DeepSeek: DeepSeek V3.1

memory 33k
stable $0.150 / $0.750 /1M

deepseek-v3.1-terminus

DeepSeek: DeepSeek V3.1 Terminus

memory 164k
stable $0.210 / $0.790 /1M

deepseek-v3.2-exp

DeepSeek: DeepSeek V3.2 Exp

memory 164k
stable $0.270 / $0.410 /1M

deepseek-v3.2

DeepSeek: DeepSeek V3.2

memory 164k
stable $0.260 / $0.380 /1M

deepseek-v3.2-speciale

DeepSeek: DeepSeek V3.2 Speciale

memory 164k
stable $0.400 / $1.20 /1M

deepseek-r1-0528

DeepSeek-R1-0528

Verbesserte R1-Version mit weniger Halluzinationen und JSON/Function Calling.

stable $0.450 / $2.15 /1M

deepseek-v3-2-speciale

DeepSeek-V3.2-Speciale

Reasoning-stärkere API-only-Variante von V3.2.

stable Zur Dokumentation arrow_forward

deepseek-v3

DeepSeek-V3

Großes generelles Basismodell der V3-Reihe.

stable $0.135 / $0.500 /1M

deepseek-r1

DeepSeek-R1

Reasoning-Modell, das DeepSeek Anfang 2025 vorgestellt hat.

stable $0.700 / $2.50 /1M

deepseek-v3-2

DeepSeek-V3.2

Aktueller V3.2-Nachfolger mit stärkerem Agentic-Fokus.

stable Zur Dokumentation arrow_forward

deepseek-v3-2-exp

DeepSeek-V3.2-Exp

Math 1620 Reasoning 1555 Chat 1300

Experimentelle Zwischenversion mit Sparse Attention.

preview Zur Dokumentation arrow_forward