Large Language Models

Amazon's cost-efficient multimodal model in the Nova family. Processes text, images, video, and documents at very low la...

Amazon Nova Micro is a text-only large language model optimized for speed and cost efficiency, available through Amazon ...

Amazon's most capable multimodal model in the Nova family. Handles complex tasks requiring deep reasoning, long document...

AWS's multimodal model available via Bedrock. Optimized for enterprise workflows with strong document understanding.

Fastest Anthropic model. Ideal as fallback LLM or for high-throughput classification and extraction tasks.

Anthropic's most capable model. Improved software engineering, complex long-running tasks, and higher-resolution vision....

Anthropic's balanced model. Full upgrade across coding, computer use, long-context reasoning, agent planning, and design...

Mistral's code-specialized model. Trained on 80+ programming languages. Top-tier for code generation and review.

Efficient enterprise model optimized for RAG with built-in citation generation and tool use.

Enterprise-focused model with built-in RAG and tool use. Optimized for business workflows and grounded generation.

Command A is Cohere's enterprise-focused large language model optimized for agentic AI, multilingual tasks, and RAG with...

Databricks' open MoE model (132B total, 36B active). Strong at structured data tasks and SQL generation.

Open reasoning model competitive with o1. Chain-of-thought reasoning with transparent thinking process.

Large language model with 671B MoE parameters optimized for code generation, reasoning, and general-purpose tasks. Uses ...

DeepSeek's latest flagship (December 2025). First open model to integrate thinking into tool use. Gold-medal performance...

Open LLM family by the Technology Innovation Institute (TII). Sizes 1B-10B trained on 14T tokens with high-quality data ...

Google's fast 2.5-gen model. Superseded by Gemini 3 Flash (2026). Great for latency-sensitive applications with strong m...

Google's powerful 2.5-gen model with 1M+ context. Superseded by Gemini 3.1 Pro (Feb 2026). Still excellent for long docu...

Google's fast next-generation model (2026). Major upgrade over Gemini 2.5 Flash. New default in Gemini app. 1M token con...

Google's most advanced reasoning model (Feb 2026). 1M token context. Solves complex problems from text, audio, images, v...

Google's on-device model. Runs locally on mobile/desktop without cloud. Built into Chrome and Android.

Google's open model (2B-27B). Superseded by Gemma 3 (2025) and Gemma 4 (2026). Still stable for on-device and edge deplo...

Google's open model series (March 2025). Sizes 1B-27B, 128K context, multimodal. Outperforms Llama 3.3 70B on several be...

Google's latest open model (April 2026). 31B params, ranked #3 open model on Arena AI. Multimodal (text, image, video, a...

GPT-4.1 is OpenAI's flagship large language model optimized for coding, instruction following, and long-context tasks. I...

Fast and affordable model from OpenAI. 82% cheaper than GPT-4o. Best for high-volume tasks where speed matters.

xAI's flagship model with real-time web and X/Twitter data (Feb 2025). Strong reasoning and coding. Grok 4 (reasoning-fi...

xAI's reasoning-first model (mid-2025). Always-on thinking for deeper analysis. Stronger than Grok 3 on benchmarks. Real...

IBM Granite 3.3 is a family of open-source language models optimized for enterprise workloads including RAG, agentic wor...

Conversational AI model focused on empathetic, helpful dialogue. Available via enterprise API for customer-facing apps.

AI21's hybrid SSM-Transformer model. 256K context window with efficient long-context processing.

Meta's proven open model. Strong all-around performance, widely available on hosting platforms.

Meta's open-weight mixture-of-experts model. 400B total params, 17 active experts. Strong multilingual and coding perfor...

Meta's efficient open model. 109B params with 16 experts. 10M token context window for massive document processing.

Desktop app for running local LLMs. User-friendly GUI, GGUF model support, built-in chat and API server.

Mistral's flagship model. Excellent for complex reasoning with strong multilingual support, especially European language...

Mistral's flagship MoE model (December 2025). 41B active params, 675B total. State-of-the-art on coding, reasoning, and ...

Mistral's reasoning model family (May 2025). Magistral Medium (24B) and Small (8B) use chain-of-thought for math, coding...

Mistral's 8B edge and on-device model. Best-in-class performance in the sub-10B range on knowledge, commonsense, and STE...

12B open model co-developed with NVIDIA. Apache 2.0 license, 128K context. Great for self-hosting.

Mistral's frontier multimodal model (124B). Processes text and images natively — reads charts, diagrams, and multi-page ...

Affordable Mistral model optimized for low-latency tasks. Great for classification, summarization, and translation.

Mistral's versatile hybrid model (March 2026). 119B params, 256K context. Unifies instruct, deep reasoning, and multimod...

NVIDIA's model optimized for enterprise. Strong at structured output and function calling for business automation.

Run LLMs locally with one command. Supports Llama, Mistral, Gemma, Phi and more. Zero cloud dependency.

Smaller, faster variant of GPT-4.1 (April 2025). Balances capability and cost for high-volume tasks. 1M token context, m...

Fastest and most cost-efficient GPT-4.1 variant (April 2025). Optimised for latency-sensitive, high-volume applications....

GPT-4o is OpenAI's multimodal large language model that accepts text, image, and audio inputs and produces text and audi...

OpenAI's flagship frontier model, released August 2025. Major intelligence leap over GPT-4o. Excels at coding, reasoning...

OpenAI's current flagship (March 2026). State-of-the-art on coding, reasoning, and professional tasks. 272K context wind...

Cost-efficient GPT-5.4 variant (March 2026). Balances speed and capability for high-volume workloads. Faster and cheaper...

Reasoning model that thinks before answering. Excels at math, science, and complex multi-step problems.

Most capable reasoning model from OpenAI. State-of-the-art on coding and math benchmarks.

Lightweight reasoning model. Good balance of reasoning capability and cost for everyday tasks.

OpenAI's highest-capability reasoning model (June 2025). Uses significantly more compute per query for the best possible...

Compact reasoning model from OpenAI (April 2025). Retired from ChatGPT in Feb 2026; still available via API. Cost-effect...

Search-augmented LLM with real-time web access. Returns cited answers grounded in current web sources.

Microsoft's small language model (14B params). Punches above its weight on reasoning. Great for on-device deployment.

Microsoft's efficient small model (3.8B params). Strong math, reasoning, and multilingual support with function calling....

Alibaba's open model from the Qwen 2.5 series. Superseded by Qwen3 (2025). Still strong at math, coding, and Chinese-Eng...

Alibaba's code-specialized model (Qwen 2.5 series). Superseded by Qwen3-Coder. Available in 7B-32B sizes. Apache 2.0.

Alibaba's major open model generation (May 2025). MoE 235B/22B active and dense variants up to 32B. Hybrid thinking mode...

Alibaba's latest flagship (April 2026). Tops six major coding benchmarks. 256K context. Qwen3.6-Plus optimized for agent...

Multimodal model handling text, images, and video. Strong at visual reasoning and document understanding.

Custom hardware-accelerated inference. Fastest token generation speeds for large models. Enterprise-focused.

Stability AI's text model family. Lightweight models (3B-12B) good for edge deployment and fine-tuning.

Korean AI company's LLM. Strong at document understanding and structured data extraction.

Enterprise LLM focused on brand-consistent content generation. Built-in compliance and style guide enforcement.

01.AI's flagship model. Strong multilingual performance with competitive reasoning on benchmarks.