Reasoning
Claude Opus 5.5
Anthropic
Anthropic's recommended starting point for most serious work: long-running agentic coding and knowledge work, with a 1M token context window.
- Anthropic
- Vision
- Tool calling
AI models
AI models for business, compared in plain English
What each model is good at, where it falls short and what to check before you build on it. Relative tiers, not benchmark scores, and a reviewed date on every page.
- 45 models reviewed
- 8 categories
Featured
Models we use most often on client projects
General purpose
Claude Sonnet 5
Anthropic
Anthropic's balance of speed and intelligence: a strong everyday model for assistants, document work, tool calling and coding.
- Anthropic
- Vision
- Tool calling
General purpose
Claude Haiku 4.5
Anthropic
Anthropic's fastest and lowest-cost Claude model, with near-frontier intelligence for high-volume and real-time work.
- Anthropic
- Vision
- Tool calling
Reasoning
Gemini 3.1 Pro
Google's most advanced Gemini model for reasoning, software engineering and agent work, reading text, images, audio, video and PDFs. Available as a preview.
- Vision
- Tool calling
General purpose
Gemini 3.8 Flash
Google's most capable Flash model, stable since September 2026, for agents, software engineering and enterprise workflows with full multimodal input.
- Vision
- Tool calling
General purpose
GPT-6 Sol
OpenAI
The middle model of the GPT-6 family, positioned for complex coding and agentic workflows at a mid price level.
- OpenAI
- Vision
- Tool calling
General purpose
GPT-6 Luna
OpenAI
OpenAI's most efficient model for focused, high-volume tasks, with vision, tool calling and structured outputs at the lowest price level in the family.
- OpenAI
- Vision
- Tool calling
Directory
All models
45 of 45 models
-
Reasoning
Claude Fable 5.1
Anthropic
The top tier of the Claude family, for the most demanding reasoning and long-horizon agent work, priced above Opus.
- Anthropic
- Vision
- Tool calling
-
Reasoning
Claude Opus 5.5
Anthropic
Anthropic's recommended starting point for most serious work: long-running agentic coding and knowledge work, with a 1M token context window.
- Anthropic
- Vision
- Tool calling
-
General purpose
Claude Sonnet 5
Anthropic
Anthropic's balance of speed and intelligence: a strong everyday model for assistants, document work, tool calling and coding.
- Anthropic
- Vision
- Tool calling
-
General purpose
Claude Haiku 4.5
Anthropic
Anthropic's fastest and lowest-cost Claude model, with near-frontier intelligence for high-volume and real-time work.
- Anthropic
- Vision
- Tool calling
-
Reasoning
Gemini 3.1 Pro
Google
Google's most advanced Gemini model for reasoning, software engineering and agent work, reading text, images, audio, video and PDFs. Available as a preview.
- Vision
- Tool calling
-
General purpose
Gemini 3.8 Flash
Google
Google's most capable Flash model, stable since September 2026, for agents, software engineering and enterprise workflows with full multimodal input.
- Vision
- Tool calling
-
General purpose
Gemini 3.5 Flash-Lite
Google
Google's lowest-cost current Gemini model for high-throughput work such as sub-agent tasks and document parsing.
- Vision
- Tool calling
-
Open weights
Gemma 4
Google
Google's open-weight model family under Apache 2.0, in sizes from phone-friendly to 31B, with image input and function calling.
- Open weights
- Vision
- Tool calling
-
Embeddings
Gemini Embedding 2
Google
Google's current embedding model, multimodal: it embeds text, images, video, audio and PDFs for search and retrieval.
- Vision
-
Image and video
Nano Banana (Gemini image)
Google
Google's Gemini-native image generation and editing models, known as Nano Banana 2 and Nano Banana Pro, which replace Imagen in the Gemini API.
- Vision
-
Image and video
Veo 3.1
Google
Google's video generation model, creating short video with native audio from text and images. Available as a preview, with fast and lite versions.
- Vision
-
Speech
Deepgram Nova-3 and Flux
Deepgram
Deepgram's speech-to-text models: Nova-3 for general transcription in 60+ languages, and Flux for voice agents with built-in end-of-turn detection.
- Deepgram
-
Speech
ElevenLabs Eleven v3
ElevenLabs
ElevenLabs text-to-speech: the expressive Eleven v3 in 70+ languages, a real-time conversational version, and the low-latency Flash models, with voice cloning.
- ElevenLabs
-
Image and video
Stable Diffusion 3.5
Stability AI
Stability AI's open-weight image models, in Large, Medium and Flash versions, which can be self-hosted or used through the Stability API.
- Stability AI
- Open weights
- Vision
-
Image and video
FLUX.2
Black Forest Labs
Black Forest Labs' FLUX.2 image models: API-only Pro, Flex and Max versions, plus open-weight Dev and Klein versions under different licenses.
- Black Forest Labs
- Open weights
- Vision
-
Image and video
Midjourney V8
Midjourney
Midjourney's image generation service, known for its visual style. V8.1 is the default since June 2026 and V8.2 followed in July. There is no public developer API.
- Midjourney
- Vision
-
Image and video
Runway Gen-4.5
Runway
Runway's video models: Gen-4.5 for text-to-video and image-to-video, and Aleph 2.0 for editing existing video, with a developer API.
- Runway
- Vision
-
Open weights
Llama 4 Scout
Meta
An open-weight Llama 4 model with image input and a very long context window, for teams that want to host a capable model themselves.
- Meta
- Open weights
- Vision
- Tool calling
-
Open weights
Llama 4 Maverick
Meta
The larger Llama 4 open-weight model, with image input and a 1 million token context window, for self-hosted assistants and analysis.
- Meta
- Open weights
- Vision
- Tool calling
-
Open weights
Llama 3.3 70B
Meta
An older, text-only open-weight Llama model with tool use and a 128K token window, still widely hosted and well understood.
- Meta
- Open weights
- Tool calling
-
General purpose
Meta Muse Spark
Meta
Meta's newer proprietary model family, offered through the Meta Model API, with text, image, video and PDF input, tool calling and a 1 million token window.
- Meta
- Vision
- Tool calling
-
General purpose
Mistral Large 3
Mistral AI
Mistral's open-weight general-purpose flagship under Apache 2.0, with image input, tool calling, structured outputs and a 256K token window.
- Mistral AI
- Open weights
- Vision
- Tool calling
-
Open weights
Mistral Small 4
Mistral AI
Mistral's efficient open-weight model under Apache 2.0, one hybrid model for instructions, reasoning and coding, with tool calling and a 256K window.
- Mistral AI
- Open weights
- Tool calling
-
Coding
Mistral Medium 3.5
Mistral AI
A newer open-weight Mistral model for agent and coding work, with image input, tool calling, structured outputs and a 256K window.
- Mistral AI
- Open weights
- Vision
- Tool calling
-
Coding
Codestral 25.08
Mistral AI
Mistral's low-latency coding model for code completion and generation, with fill-in-the-middle support, tool calling and a 128K window.
- Mistral AI
- Tool calling
-
Vision and multimodal
Mistral OCR 4.1
Mistral AI
Mistral's document OCR model, turning pages into structured output with bounding boxes, block labels and confidence scores.
- Mistral AI
- Vision
-
Reasoning
GPT-6 Astra
OpenAI
OpenAI's most capable model, built for the hardest end-to-end work: complex reasoning, coding, computer use and research.
- OpenAI
- Vision
- Tool calling
-
General purpose
GPT-6 Sol
OpenAI
The middle model of the GPT-6 family, positioned for complex coding and agentic workflows at a mid price level.
- OpenAI
- Vision
- Tool calling
-
General purpose
GPT-6 Luna
OpenAI
OpenAI's most efficient model for focused, high-volume tasks, with vision, tool calling and structured outputs at the lowest price level in the family.
- OpenAI
- Vision
- Tool calling
-
General purpose
GPT-4.1
OpenAI
An older OpenAI model described as its smartest non-reasoning model, strong at following instructions and calling tools, with a very long context window.
- OpenAI
- Vision
- Tool calling
-
General purpose
GPT-4o mini
OpenAI
A compact, low-cost older OpenAI model for focused tasks, with vision, tool calling and structured outputs and a 128K token window.
- OpenAI
- Vision
- Tool calling
-
Embeddings
text-embedding-3-large
OpenAI
OpenAI's most capable embedding model for search and retrieval, with 3,072-dimension vectors and support for English and other languages.
- OpenAI
-
Embeddings
text-embedding-3-small
OpenAI
OpenAI's efficient, lowest-cost embedding model, with 1,536-dimension vectors for search, retrieval and similarity at scale.
- OpenAI
-
Speech
GPT Transcribe
OpenAI
OpenAI's current speech-to-text model, replacing Whisper and the GPT-4o transcribe models, with streaming and keyword or language hints.
- OpenAI
-
Image and video
GPT Image 2.5
OpenAI
OpenAI's current image generation and editing models: Sunburst for the highest quality and Flare for fast everyday images.
- OpenAI
- Vision
-
Speech
GPT-4o mini TTS
OpenAI
OpenAI's current text-to-speech model for turning text into natural spoken audio, at a low price level.
- OpenAI
-
General purpose
DeepSeek V4.1 Flash
DeepSeek
DeepSeek's default, best-value model: open weights under MIT, image input, tool calling, JSON output, optional thinking and a 1 million token window.
- DeepSeek
- Open weights
- Vision
- Tool calling
-
Reasoning
DeepSeek V4 Pro
DeepSeek
DeepSeek's model for demanding reasoning and agent work, with thinking effort levels, tool calling, JSON output and a 1 million token window.
- DeepSeek
- Open weights
- Tool calling
-
Reasoning
Qwen3.8-Max
Alibaba Cloud
Alibaba's current Qwen flagship for long autonomous coding and professional work, with image and video input, tool calling, JSON Schema output and a 1 million token window.
- Alibaba Cloud
- Vision
- Tool calling
-
Coding
Qwen3-Coder
Alibaba Cloud
Qwen's open-weight coding models under Apache 2.0, from the efficient Qwen3-Coder-Next to the large 480B model, with tool use and long context.
- Alibaba Cloud
- Open weights
- Tool calling
-
Reasoning
Grok 4.7
xAI
xAI's current flagship for coding, agent tasks and knowledge work, with image input, tool calling, structured outputs, reasoning effort levels and a 500K window.
- xAI
- Vision
- Tool calling
-
General purpose
Cohere Command A+
Cohere
Cohere's enterprise flagship for multimodal, multilingual agent tasks, with 48 languages, tool calling, JSON output and open weights under Apache 2.0.
- Cohere
- Open weights
- Vision
- Tool calling
-
Embeddings
Cohere Embed v4
Cohere
Cohere's multimodal embedding model for enterprise search, embedding text, images and mixed documents, with flexible vector sizes and long inputs.
- Cohere
- Vision
-
Embeddings
Cohere Rerank 4
Cohere
Cohere's multilingual rerank models, which sort search results by relevance, in a best-quality Pro version and a low-latency Fast version.
- Cohere
-
Embeddings
Voyage 4
Voyage AI by MongoDB
Voyage AI's current embedding series, now part of MongoDB, with large, standard and lite models that share one embedding space.
- Voyage AI by MongoDB
Nothing matches those filters. or ask us.
Browse
Browse by category
10 models
Reasoning
Models that think through multi-step problems before answering: analysis, planning, math, complex documents and agent work.
21 models
General purpose
All-round language models for writing, summarizing, support, extraction and most everyday business tasks.
14 models
Coding
Models that write, review and explain code, and power coding assistants and developer tools.
25 models
Vision and multimodal
Models that read images, scans, screenshots and sometimes audio or video alongside text.
6 models
Embeddings
Models that turn text into vectors for search, retrieval, clustering and recommendations.
4 models
Speech
Speech-to-text and text-to-speech models for transcription, call analysis and voice assistants.
7 models
Image and video
Models that generate or edit images and video from text and reference images.
13 models
Open weights
Models whose weights you can download and run on your own servers or a cloud of your choice.
Model picker
Which model should I use?
Six quick questions give you a category and two or three models to test first. It is a starting point, not a verdict.
How to read this directory
Tiers, not scores
Each model page shows eight relative tiers from one to five: reasoning, coding, vision, speed, price level, context size, tool calling and structured output. They compare models with each other at the time of review. They are our judgment from provider documentation and our own use, not benchmark results.
Every page also lists what the model is good at, where it is not the right choice, the business tasks we would use it for and what to check before you commit. Prices and limits change often, so each page carries a reviewed date and a link to the provider documentation.
To choose for a real project, we test two or three candidates on your own examples. That is what our AI model evaluation service does, and it usually takes days, not weeks. See also LLM integration, MCP servers and the problems we solve with these models.
Keep exploring
Related on ExecMedia
Related services
View all related services- AI Product Development AI agents, knowledge assistants, copilots and document automation built into the way your team already works.
- LLM integration Add a large language model to software you already have, with the guardrails, costs and logging handled.
- AI model evaluation Test candidate AI models on your real data and tasks, then pick the one that balances quality, speed and cost.
- AI agents AI that takes actions in your systems, such as qualifying leads or processing requests, with people checking the results.
- AI copilots AI helpers built into your product or internal tools that draft, summarize and suggest while people decide.
- Voice AI Speech to text, call summaries and voice assistants that take bookings or answer routine calls.
Solutions
View all solutionsIndustries
View all industriesCase studies
View all case studiesGuides and articles
View all guides and articles- How to choose an AI model for your business A step-by-step way to pick an AI model by testing candidates on your own tasks, data rules and budget.
- How to keep customer data safe when using AI Practical steps to protect customer data when you use AI services, from data terms to access and logging.
- RAG vs fine-tuning Two ways to make AI work with your knowledge, compared by cost, accuracy and upkeep.
AI models
View all ai models- Gemma 4 Google's open-weight model family under Apache 2.0, in sizes from phone-friendly to 31B, with image input and function calling.
- Llama 4 Scout An open-weight Llama 4 model with image input and a very long context window, for teams that want to host a capable model themselves.
- Llama 4 Maverick The larger Llama 4 open-weight model, with image input and a 1 million token context window, for self-hosted assistants and analysis.
- Llama 3.3 70B An older, text-only open-weight Llama model with tool use and a 128K token window, still widely hosted and well understood.
- Meta Muse Spark Meta's newer proprietary model family, offered through the Meta Model API, with text, image, video and PDF input, tool calling and a 1 million token window.
- Mistral Small 4 Mistral's efficient open-weight model under Apache 2.0, one hybrid model for instructions, reasoning and coding, with tool calling and a 256K window.
Glossary terms
View all glossary terms- Large language model A large language model, or LLM, is an AI model trained on vast amounts of text that can understand and generate language, and often images and code.
- Context window A context window is the maximum amount of text, measured in tokens, that an AI model can consider at once, including the question, documents and its answer.
- Fine-tuning Fine-tuning is further training an existing AI model on your own examples so it learns a specific style, format or task.
- Inference Inference is the step where a trained AI model is used to produce an output, such as an answer, a label or a prediction, from new input.
- LLM integration LLM integration is connecting a large language model to your software and data, so AI features work inside your own products and processes.
- Open-weights model An open-weights model is an AI model whose trained parameters are published, so anyone can download and run it on their own hardware under its license.
FAQ
Questions people ask us
Have a question that is not here? Ask us directly.
There is no single best model. The right one depends on the task, volume, speed, budget and data rules. Use the model picker on this page for a starting point, then test two or three candidates on your own examples. We do this as part of AI model evaluation.
Benchmarks change often and rarely match business tasks. Relative tiers give a fair first impression. Real decisions should come from tests on your own data.
Every model page shows the date we last checked it against the provider documentation. Prices and limits change often, so always confirm them with the provider before you commit.
Most production systems use more than one: a fast, low-cost model for simple steps and a stronger model for the hard ones. Building with a thin layer between your code and the provider also makes switching easier later. See LLM integration.