LLMs nach Kategorien
Top KI-Modelle nach Kategorie
Vergleiche aktuelle Modelle aus Open Source, proprietär, uncensored, Coding, Mathe, Geschwindigkeit und Neuerscheinungen.
Meistgenutzte KI-Modelle
Beliebte Modelle auf OpenRouter.
MiMo-V2-Pro
Xiaomi
MiMo-V2-Pro is Xiaomi's flagship foundation model, featuring over 1T total parameters and a 1M context length, deeply optimized for agentic scenarios.
MiniMax M2.5
MiniMax
MiniMax-M2.5 is a SOTA large language model designed for real-world productivity.
MiniMax M2.7
MiniMax
MiniMax-M2.7 is a next-generation large language model designed for autonomous, real-world productivity and continuous improvement.
GPT-5.4
OpenAI
GPT-5.4 is OpenAI’s latest frontier model, unifying the Codex and GPT lines into a single system.
Gemini 3.1 Pro Preview
Gemini 3.1 Pro Preview is Google’s frontier reasoning model, delivering enhanced software engineering performance, improved agentic reliability, and more efficient token usage across complex workflows.
GLM 5.1
NEWZ.ai
GLM-5.1 delivers a major leap in coding capability, with particularly significant gains in handling long-horizon tasks.
Kimi K2.5
MoonshotAI
Kimi K2.5 is Moonshot AI's native multimodal model, delivering state-of-the-art visual coding capability and a self-directed agent swarm paradigm.
Step 3.5 Flash
StepFun
Step 3.5 Flash is StepFun's most capable open-source foundation model.
Gemini 3.1 Flash Lite Preview
Gemini 3.1 Flash Lite Preview is Google's high-efficiency model optimized for high-volume use cases.
GLM 5 Turbo
Z.ai
GLM-5 Turbo is a new model from Z.ai designed for fast inference and strong performance in agent-driven environments such as OpenClaw scenarios.
Top Open-Source-KI-Modelle
Community-getrieben und transparent.
GLM 5.1
NEWZ.ai
GLM-5.1 delivers a major leap in coding capability, with particularly significant gains in handling long-horizon tasks.
Qwen3.5 Plus
Qwen
The Qwen3.5 native vision-language series Plus models are built on a hybrid architecture that integrates linear attention mechanisms with sparse mixture-of-experts models, achieving higher inference efficiency.
MiniMax M2.7
MiniMax
MiniMax-M2.7 is a next-generation large language model designed for autonomous, real-world productivity and continuous improvement.
MiMo-V2-Pro
Xiaomi
MiMo-V2-Pro is Xiaomi's flagship foundation model, featuring over 1T total parameters and a 1M context length, deeply optimized for agentic scenarios.
Kimi K2.5
MoonshotAI
Kimi K2.5 is Moonshot AI's native multimodal model, delivering state-of-the-art visual coding capability and a self-directed agent swarm paradigm.
GLM 5 Turbo
Z.ai
GLM-5 Turbo is a new model from Z.ai designed for fast inference and strong performance in agent-driven environments such as OpenClaw scenarios.
Qwen3.5 397B A17B
Qwen
The Qwen3.5 series 397B-A17B native vision-language model is built on a hybrid architecture that integrates a linear attention mechanism with a sparse mixture-of-experts model, achieving higher inference efficiency.
MiMo-V2-Omni
Xiaomi
MiMo-V2-Omni is a frontier omni-modal model that natively processes image, video, and audio inputs within a unified architecture.
Qwen3.5-35B-A3B
Qwen
The Qwen3.5 Series 35B-A3B is a native vision-language model designed with a hybrid architecture that integrates linear attention mechanisms and a sparse mixture-of-experts model, achieving higher inference efficiency.
GLM 5V Turbo
Z.ai
GLM-5V-Turbo is Z.ai’s first native multimodal agent foundation model, built for vision-based coding and agent-driven tasks.
Top Proprietäre KI-Modelle
Führende Closed-Source-Modelle.
Claude Opus 4.6 (Fast)
NEWAnthropic
Fast-mode variant of [Opus 4.6](/anthropic/claude-opus-4.6) - identical capabilities with higher output speed at premium 6x pricing.
Gemini 3.1 Pro Preview
Gemini 3.1 Pro Preview is Google’s frontier reasoning model, delivering enhanced software engineering performance, improved agentic reliability, and more efficient token usage across complex workflows.
GPT-5.4
OpenAI
GPT-5.4 is OpenAI’s latest frontier model, unifying the Codex and GPT lines into a single system.
GPT-5.3-Codex
OpenAI
GPT-5.3-Codex is OpenAI’s most advanced agentic coding model, combining the frontier software engineering performance of GPT-5.2-Codex with the broader reasoning and professional knowledge capabilities of GPT-5.2.
Grok 4.20
xAI
Grok 4.20 is xAI's newest flagship model with industry-leading speed and agentic tool calling capabilities.
GPT-5.4 Mini
OpenAI
GPT-5.4 mini brings the core capabilities of GPT-5.4 to a faster, more efficient model optimized for high-throughput workloads.
GPT-5.1-Codex-Max
OpenAI
GPT-5.1-Codex-Max is OpenAI’s latest agentic coding model, designed for long-running, high-context software development tasks.
Gemini 3.1 Flash Lite Preview
Gemini 3.1 Flash Lite Preview is Google's high-efficiency model optimized for high-volume use cases.
GPT-5.1-Codex-Mini
OpenAI
GPT-5.1-Codex-Mini is a smaller and faster version of GPT-5.1-Codex
Gemma 4 26B A4B
Gemma 4 26B A4B IT is an instruction-tuned Mixture-of-Experts (MoE) model from Google DeepMind.
Top Coding-KI-Modelle
Optimiert für Code und Entwickler-Workflows.
GPT-5.4
OpenAI
GPT-5.4 is OpenAI’s latest frontier model, unifying the Codex and GPT lines into a single system.
Gemini 3.1 Pro Preview
Gemini 3.1 Pro Preview is Google’s frontier reasoning model, delivering enhanced software engineering performance, improved agentic reliability, and more efficient token usage across complex workflows.
GPT-5.3-Codex
OpenAI
GPT-5.3-Codex is OpenAI’s most advanced agentic coding model, combining the frontier software engineering performance of GPT-5.2-Codex with the broader reasoning and professional knowledge capabilities of GPT-5.2.
Claude Opus 4.6 (Fast)
NEWAnthropic
Fast-mode variant of [Opus 4.6](/anthropic/claude-opus-4.6) - identical capabilities with higher output speed at premium 6x pricing.
GPT-5.4 Mini
OpenAI
GPT-5.4 mini brings the core capabilities of GPT-5.4 to a faster, more efficient model optimized for high-throughput workloads.
GPT-5.1-Codex-Max
OpenAI
GPT-5.1-Codex-Max is OpenAI’s latest agentic coding model, designed for long-running, high-context software development tasks.
GLM 5.1
NEWZ.ai
GLM-5.1 delivers a major leap in coding capability, with particularly significant gains in handling long-horizon tasks.
Qwen3.5 Plus
Qwen
The Qwen3.5 native vision-language series Plus models are built on a hybrid architecture that integrates linear attention mechanisms with sparse mixture-of-experts models, achieving higher inference efficiency.
Gemini 3.1 Flash Lite Preview
Gemini 3.1 Flash Lite Preview is Google's high-efficiency model optimized for high-volume use cases.
Grok 4.20
xAI
Grok 4.20 is xAI's newest flagship model with industry-leading speed and agentic tool calling capabilities.
Top OCR-KI-Modelle
Spezialisiert auf Texterkennung und Dokument-Extraktion.
PaddleOCR-VL-0.9B
PaddlePaddle
Baidu's 0.9B vision-language OCR model combining a NaViT-style dynamic-resolution encoder with ERNIE-4.5-0.3B. Handles multilingual text, tables, charts, and formulas across 16K context — optimized for efficient on-device document parsing.
olmOCR-2-7B
AllenAI
Allen AI's 7B OCR model fine-tuned from Qwen2.5-VL-7B on curated academic papers and technical documentation. Supports 128K context and extracts structured text from PDFs and scanned documents with high fidelity.
DeepSeek-OCR
DeepSeek
DeepSeek's ~3B MoE OCR model using optical context compression to encode full pages into compact token sequences. Outputs structured Markdown preserving text layout, tables, and mathematical formulas from images and PDFs.
Mistral OCR
Mistral AI
Mistral's dedicated document understanding model (December 2025). Processes PDFs and images page-by-page via API, returning structured Markdown with preserved tables, equations, image bounding boxes, and rich layout metadata.
Top Mathe-KI-Modelle
Spezialisten für Mathematik und Reasoning.
GPT-5.4
OpenAI
GPT-5.4 is OpenAI’s latest frontier model, unifying the Codex and GPT lines into a single system.
GPT-5.3-Codex
OpenAI
GPT-5.3-Codex is OpenAI’s most advanced agentic coding model, combining the frontier software engineering performance of GPT-5.2-Codex with the broader reasoning and professional knowledge capabilities of GPT-5.2.
Gemini 3.1 Flash Lite Preview
Gemini 3.1 Flash Lite Preview is Google's high-efficiency model optimized for high-volume use cases.
DeepSeek V3.2 Speciale
DeepSeek
DeepSeek-V3.2-Speciale is a high-compute variant of DeepSeek-V3.2 optimized for maximum reasoning and agentic performance.
MiMo-V2-Flash
Xiaomi
MiMo-V2-Flash is an open-source foundation language model developed by Xiaomi.
GPT-5.1-Codex-Mini
OpenAI
GPT-5.1-Codex-Mini is a smaller and faster version of GPT-5.1-Codex
Gemini 3.1 Pro Preview
Gemini 3.1 Pro Preview is Google’s frontier reasoning model, delivering enhanced software engineering performance, improved agentic reliability, and more efficient token usage across complex workflows.
GLM 4 32B
Z.ai
GLM 4 32B is a cost-effective foundation language model.
Kimi K2 Thinking
MoonshotAI
Kimi K2 Thinking is Moonshot AI’s most advanced open reasoning model to date, extending the K2 series into agentic, long-horizon reasoning.
GPT-5.1-Codex-Max
OpenAI
GPT-5.1-Codex-Max is OpenAI’s latest agentic coding model, designed for long-running, high-context software development tasks.
Schnelle KI-Modelle
Niedrige Kosten bei geringer Latenz.
Mercury 2
Inception
Mercury 2 is an extremely fast reasoning LLM, and the first reasoning diffusion LLM (dLLM).
Gemini 3.1 Flash Lite Preview
Gemini 3.1 Flash Lite Preview is Google's high-efficiency model optimized for high-volume use cases.
gpt-oss-safeguard-20b
OpenAI
gpt-oss-safeguard-20b is a safety reasoning model from OpenAI built upon gpt-oss-20b.
Ministral 3 3B 2512
Mistral
The smallest model in the Ministral 3 family, Ministral 3 3B is a powerful, efficient tiny language model with vision capabilities.
Qwen3.5-Flash
Qwen
The Qwen3.5 native vision-language Flash models are built on a hybrid architecture that integrates a linear attention mechanism with a sparse mixture-of-experts model, achieving higher inference efficiency.
Qwen3.5-27B
Qwen
The Qwen3.5 27B native vision-language Dense model incorporates a linear attention mechanism, delivering fast response times while balancing inference speed and performance.
Qwen3.5-35B-A3B
Qwen
The Qwen3.5 Series 35B-A3B is a native vision-language model designed with a hybrid architecture that integrates linear attention mechanisms and a sparse mixture-of-experts model, achieving higher inference efficiency.
Grok 4.20
xAI
Grok 4.20 is xAI's newest flagship model with industry-leading speed and agentic tool calling capabilities.
Mistral Small 4
Mistral
Mistral Small 4 is the next major release in the Mistral Small family, unifying the capabilities of several flagship Mistral models into a single system.
GPT-5.1-Codex-Mini
OpenAI
GPT-5.1-Codex-Mini is a smaller and faster version of GPT-5.1-Codex
Top Bildgenerierungs-KI-Modelle
Modelle, die aus Textprompts Bilder erzeugen.
Nano Banana 2 (Gemini 3.1 Flash Image Preview)
Gemini 3.1 Flash Image Preview, a.k.a.
Seedream 4.5
ByteDance Seed
Seedream 4.5 is the latest in-house image generation model developed by ByteDance.
Riverflow V2 Fast
Sourceful
Riverflow V2 Fast is the fastest variant of Sourceful's Riverflow 2.0 lineup, best for production deployments and latency-critical workflows.
Riverflow V2 Pro
Sourceful
Riverflow V2 Pro is the most powerful variant of Sourceful's Riverflow 2.0 lineup, best for top-tier control and perfect text rendering.
Riverflow V2 Standard Preview
Sourceful
Riverflow V2 Standard Preview is the standard variant of Sourceful's Riverflow V2 preview lineup.
FLUX.2 Pro
Black Forest Labs
A high-end image generation and editing model focused on frontier-level visual quality and reliability.
FLUX.2 Max
Black Forest Labs
FLUX.2 [max] is the new top-tier image model from Black Forest Labs, pushing image quality, prompt understanding, and editing consistency to the highest level yet.
Riverflow V2 Max Preview
Sourceful
Riverflow V2 Max Preview is the most powerful variant of Sourceful's Riverflow V2 preview lineup.
FLUX.2 Klein 4B
Black Forest Labs
FLUX.2 [klein] 4B is the fastest and most cost-effective model in the FLUX.2 family, optimized for high-throughput use cases while maintaining excellent image quality.
FLUX.2 Flex
Black Forest Labs
FLUX.2 [flex] excels at rendering complex text, typography, and fine details, and supports multi-reference editing in the same unified architecture.
Top Audio-KI-Modelle
Modelle mit Sprach- und Audio-Ausgabe.
GPT Audio Mini
OpenAI
A cost-efficient version of GPT Audio.
GPT Audio
OpenAI
The gpt-audio model is OpenAI's first generally available audio model.
KI-Modelle mit großem Kontextfenster
Modelle mit 200K+ Kontextfenstern.
Grok 4.20
xAI
Grok 4.20 is xAI's newest flagship model with industry-leading speed and agentic tool calling capabilities.
Grok 4.20 Multi-Agent
xAI
Grok 4.20 Multi-Agent is a variant of xAI’s Grok 4.20 designed for collaborative, agent-based workflows.
GPT-5.4
OpenAI
GPT-5.4 is OpenAI’s latest frontier model, unifying the Codex and GPT lines into a single system.
GPT-5.4 Pro
OpenAI
GPT-5.4 Pro is OpenAI's most advanced model, building on GPT-5.4's unified architecture with enhanced reasoning capabilities for complex, high-stakes tasks.
MiMo-V2-Pro
Xiaomi
MiMo-V2-Pro is Xiaomi's flagship foundation model, featuring over 1T total parameters and a 1M context length, deeply optimized for agentic scenarios.
Gemini 3.1 Pro Preview
Gemini 3.1 Pro Preview is Google’s frontier reasoning model, delivering enhanced software engineering performance, improved agentic reliability, and more efficient token usage across complex workflows.
Gemini 3.1 Flash Lite Preview
Gemini 3.1 Flash Lite Preview is Google's high-efficiency model optimized for high-volume use cases.
Gemini 3.1 Pro Preview Custom Tools
Gemini 3.1 Pro Preview Custom Tools is a variant of Gemini 3.1 Pro that improves tool selection behavior by preventing overuse of a general bash tool when more efficient third-party or user-defined functions are available.
Palmyra X5
Writer
Palmyra X5 is Writer's most advanced model, purpose-built for building and scaling AI agents across the enterprise.
Qwen3.5-Flash
Qwen
The Qwen3.5 native vision-language Flash models are built on a hybrid architecture that integrates a linear attention mechanism with a sparse mixture-of-experts model, achieving higher inference efficiency.
Top Uncensored KI-Modelle
Leicht gefilterte Modelle mit hoher Flexibilität.
Neueste KI-Modelle
Frische Releases auf OpenRouter.
Claude Opus 4.6 (Fast)
NEWAnthropic
Fast-mode variant of [Opus 4.6](/anthropic/claude-opus-4.6) - identical capabilities with higher output speed at premium 6x pricing.
GLM 5.1
NEWZ.ai
GLM-5.1 delivers a major leap in coding capability, with particularly significant gains in handling long-horizon tasks.
Gemma 4 26B A4B
Gemma 4 26B A4B IT is an instruction-tuned Mixture-of-Experts (MoE) model from Google DeepMind.
GLM 5V Turbo
Z.ai
GLM-5V-Turbo is Z.ai’s first native multimodal agent foundation model, built for vision-based coding and agent-driven tasks.
Trinity Large Thinking
Arcee AI
Trinity Large Thinking is a powerful open source reasoning model from the team at Arcee AI.
Grok 4.20 Multi-Agent
xAI
Grok 4.20 Multi-Agent is a variant of xAI’s Grok 4.20 designed for collaborative, agent-based workflows.
Grok 4.20
xAI
Grok 4.20 is xAI's newest flagship model with industry-leading speed and agentic tool calling capabilities.
KAT-Coder-Pro V2
Kwaipilot
KAT-Coder-Pro V2 is the latest high-performance model in KwaiKAT’s KAT-Coder series, designed for complex enterprise-grade software engineering and SaaS integration.
Reka Edge
rekaai
Reka Edge is an extremely efficient 7B multimodal vision-language model that accepts image/video+text inputs and generates text outputs.
MiMo-V2-Omni
Xiaomi
MiMo-V2-Omni is a frontier omni-modal model that natively processes image, video, and audio inputs within a unified architecture.
Made in Europe Chat with 100+ AI Models in one App.
Use Claude, ChatGPT, Gemini alongside with EU-Hosted Models like Deepseek, GLM-5, Kimi K2.5 and many more.
So findest du das richtige KI-Modell
Ein praktischer Leitfaden für die Modellwahl nach Use Case.
Modell an Aufgabe anpassen
Allzweckmodelle wie GPT-4o und Claude Sonnet funktionieren für viele Aufgaben gut. Für Spezialfälle sind Coding-Modelle (DeepSeek Coder, Codestral) und Mathe-Modelle (QwQ, DeepSeek R1) oft präziser und günstiger pro Token.
Kontextfenster berücksichtigen
Für lange Dokumente, Codebasen oder lange Konversationen ist die Kontextgröße entscheidend. Modelle reichen von 8K bis über 1M Token. Größere Fenster erlauben mehr Input, erhöhen aber häufig Kosten und Latenz.
Kosten, Geschwindigkeit und Qualität balancieren
Frontier-Modelle liefern Top-Benchmarkwerte, sind aber teurer und oft langsamer. Schnelle Modelle wie Gemini Flash, Llama 3 (8B) und Mistral Small sind für hohe Last oft deutlich günstiger und reagieren schneller.
Open Source vs. proprietär
Open-Source-Modelle (Llama, Mistral, Qwen, DeepSeek) bieten Self-Hosting und Anpassbarkeit. Proprietäre Modelle (GPT-4o, Claude, Gemini) führen häufig bei Benchmarks. Viele Teams kombinieren beide Ansätze.
Multimodalität prüfen
Einige Modelle verarbeiten neben Text auch Bilder, Audio oder Dateien. Für Workflows mit Screenshots, Diagrammen oder Audio-Transkripten sind Vision-/Audio-Inputs wichtig. Strukturierte Outputs und Function Calling sind zentral für Agenten.
Benchmarks als Ausgangspunkt nutzen
GPQA, MMLU Pro und HLE messen Wissen und Reasoning. LiveCodeBench und SciCode testen Coding-Praxis. MATH 500 und AIME bewerten mathematisches Lösen. Vergleiche relevante Kategorien und teste zusätzlich mit eigenen Prompts.
Daten auf dieser Seite stammen von OpenRouter und Artificial Analysis. Preise, Geschwindigkeit und Benchmark-Scores werden regelmäßig aktualisiert. Du kannst jedes Modell sofort im kostenlosen Chat testen.