# AIblogly > AI model rankings, published benchmark scores, API prices and side-by-side comparisons for text, image, video and audio models. Every figure is taken from a named source - the model's vendor, LMArena or Epoch AI - and carries the date it was read. The site publishes no scores of its own. ## Rankings and benchmarks - [AI model rankings and benchmarks](https://www.aiblogly.com/benchmarks): text and coding models ranked by LMArena Elo, with Agent Arena, Epoch AI benchmark results (GPQA Diamond, FrontierMath, Mock AIME, SimpleQA Verified and others), speed and price. - [All AI models](https://www.aiblogly.com/models): one page per model with its vendor-published price, context window, release date and benchmark figures. ## Data - [models.json](https://www.aiblogly.com/models.json): every published model as JSON - vendor, category, release date, context window, input/output price per million tokens, licence, LMArena Elo with its retrieval date, and the page that cites the source. - [MCP server](https://www.aiblogly.com/mcp) (install instructions: https://www.aiblogly.com/mcp-server): Model Context Protocol endpoint (Streamable HTTP, no auth, read-only) with four tools - search_models, get_model, compare_models, top_models - returning the same figures as the model pages, each with the page URL. ## Comparisons - [AI model comparisons](https://www.aiblogly.com/comparisons): side-by-side comparison tools and category comparisons for coding, image, video and audio models. - [Image generation models](https://www.aiblogly.com/comparisons/image-generation) - [Video generation models](https://www.aiblogly.com/comparisons/video-generation) - [Audio generation models](https://www.aiblogly.com/comparisons/audio-generation) ## News and learning - [AI news](https://www.aiblogly.com/news): model launches and announcements, each linked to the vendor's own post. - [Guides](https://www.aiblogly.com/guides): explainers on AI concepts, each opening with a short answer and ending with an FAQ. - [How LLMs work](https://www.aiblogly.com/articles/how-llms-work): tokens, embeddings, attention, training, context windows and hallucination, for beginners. - [Mixture of experts (MoE)](https://www.aiblogly.com/articles/mixture-of-experts-moe-explained): experts, the router, total vs active parameters, with published counts from vendors' model cards. - [Diffusion models](https://www.aiblogly.com/articles/diffusion-models-explained): how image and video generators turn noise into pictures, latent diffusion, guidance and flow matching, from the original papers. - [What is MCP?](https://www.aiblogly.com/articles/model-context-protocol-mcp-explained): the Model Context Protocol - hosts, clients, servers and connectors. - [What is a model card?](https://www.aiblogly.com/articles/how-to-read-an-ai-model-card) - [How much does Claude Opus cost?](https://www.aiblogly.com/articles/claude-opus-5-explained): API price of every Opus model, from Anthropic's pricing page. - Basics: [What is AI?](https://www.aiblogly.com/articles/what-is-ai), [Machine learning basics](https://www.aiblogly.com/articles/machine-learning-basics), [Computer vision](https://www.aiblogly.com/articles/computer-vision-explained), [Tokens and context windows](https://www.aiblogly.com/articles/tokens-and-context-windows-explained), [Getting started with AI tools](https://www.aiblogly.com/articles/getting-started-ai-tools), [How to compare AI models](https://www.aiblogly.com/articles/how-to-compare-ai-models), [Running models locally](https://www.aiblogly.com/articles/run-open-models-locally-llama-cpp-ollama-vllm), [Open weights vs open source](https://www.aiblogly.com/articles/open-weight-ai-models-explained), [The open-source software behind AIblogly](https://www.aiblogly.com/articles/open-source-stack-behind-aiblogly) - Prompting: [Prompt engineering](https://www.aiblogly.com/articles/prompt-engineering-guide), [Few-shot prompting](https://www.aiblogly.com/articles/few-shot-prompting-in-context-learning), [Chain-of-thought prompting](https://www.aiblogly.com/articles/chain-of-thought-prompting-explained) - Research: [Landmark AI papers](https://www.aiblogly.com/articles/ai-research-papers-2026), [Attention Is All You Need](https://www.aiblogly.com/articles/attention-is-all-you-need-transformer-paper), [Scaling laws](https://www.aiblogly.com/articles/scaling-laws-kaplan-chinchilla-explained), [Vision Transformer](https://www.aiblogly.com/articles/vision-transformer-vit-explained), [Segment Anything](https://www.aiblogly.com/articles/segment-anything-sam-explained) - Ethics and regulation: [AI ethics](https://www.aiblogly.com/articles/ai-ethics-guide), [EU AI Act](https://www.aiblogly.com/articles/eu-ai-act-timeline-risk-tiers), [NIST AI RMF](https://www.aiblogly.com/articles/nist-ai-risk-management-framework-explained) - Models: [Claude Fable 5.1](https://www.aiblogly.com/articles/claude-fable-5-1-explained), [Claude Fable 5](https://www.aiblogly.com/articles/claude-fable-5-explained), [Claude Sonnet 5](https://www.aiblogly.com/articles/claude-sonnet-5-explained), [Claude Opus 4.8](https://www.aiblogly.com/articles/claude-opus-4-8-explained), [GPT-6 Astra](https://www.aiblogly.com/articles/gpt-6-astra-explained), [GPT-5.6](https://www.aiblogly.com/articles/gpt-5-6-explained), [Gemini 4 Argon](https://www.aiblogly.com/articles/gemini-4-argon-explained), [Gemini 3.8 Live](https://www.aiblogly.com/articles/gemini-3-8-live-explained), [Gemini 3.1 Pro](https://www.aiblogly.com/articles/gemini-3-1-pro-explained), [Gemini 3.7 Flash](https://www.aiblogly.com/articles/gemini-3-7-flash-explained), [Grok 4.6](https://www.aiblogly.com/articles/grok-4-6-explained), [Grok 4.5](https://www.aiblogly.com/articles/grok-4-5-explained), [DeepSeek V4-Pro](https://www.aiblogly.com/articles/deepseek-v4-pro-explained), [DeepSeek V4-Flash](https://www.aiblogly.com/articles/deepseek-v4-flash-explained), [GLM-5.3](https://www.aiblogly.com/articles/glm-5-3-explained), [Qwen3.8-Max](https://www.aiblogly.com/articles/qwen-3-8-max-explained), [Qwen3.7 Max](https://www.aiblogly.com/articles/qwen-3-7-max-explained), [Muse Spark 1.2](https://www.aiblogly.com/articles/muse-spark-1-2-explained), [Nemotron 3.5 Lightning](https://www.aiblogly.com/articles/nemotron-3-5-lightning-explained) - [Free AI courses](https://www.aiblogly.com/free-ai-courses): free courses and videos with level, length, stated prerequisites and certificates, and each provider's own published view or sign-up figures where available. One page per topic: - [AI for beginners](https://www.aiblogly.com/free-ai-courses/ai-for-beginners) - [Prompt engineering](https://www.aiblogly.com/free-ai-courses/prompt-engineering) - [Large language models](https://www.aiblogly.com/free-ai-courses/large-language-models) - [AI engineering](https://www.aiblogly.com/free-ai-courses/ai-engineering) - [AI agents and MCP](https://www.aiblogly.com/free-ai-courses/ai-agents) - [Claude](https://www.aiblogly.com/free-ai-courses/claude) - [AI tools](https://www.aiblogly.com/tools) - [Resources](https://www.aiblogly.com/resources): curated links - MCP servers, research papers, datasets, courses, communities. ## Sources - LMArena leaderboard dataset (CC-BY-4.0): https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset - Epoch AI benchmark data (CC-BY): https://epoch.ai/benchmarks - Model prices and specifications: each vendor's own documentation, linked from the model's page. ## About - [About](https://www.aiblogly.com/about) - [Contact](https://www.aiblogly.com/contact) - [Sitemap](https://www.aiblogly.com/sitemap.xml)