Explore AI models with community-written guides. Learn what each model excels at and how to craft the best prompts for it.
Submit a modelAnthropic
Anthropic's most powerful model — exceptional at analysis, writing, and code generation.
Anthropic
Balanced performance model — fast, capable, and cost-effective for most tasks.
Anthropic
Fastest Claude model with near-Sonnet quality. Extended thinking, computer use, image understanding. One-third the cost of Sonnet.
Anthropic
Near-Opus performance at Sonnet pricing. Anthropic's best balance of intelligence, speed, and cost for production use.
Cohere
DeepSeek
Open-source reasoning model rivaling frontier closed models. Exceptional at math, code, and step-by-step reasoning.
OpenAI
OpenAI's stable API alias `chat-latest` that always resolves to the latest Instant chat model used in ChatGPT. Auto-tracks new Instant model rollouts so apps stay current without code changes. 400K context.
OpenAI
Multimodal model excelling at reasoning, coding, and natural conversation.
OpenAI
OpenAI's most capable model with advanced reasoning and tool use.
OpenAI
Latest GPT-5 variant with 1.05M token context, 33% fewer factual errors. Standard, Thinking, and Pro variants available.
Google's flagship model with native multimodal reasoning and massive context window.
Google's GA high-efficiency multimodal model optimized for low-latency, high-volume workloads. Supports text, image, video, audio, and PDF inputs. Designed for lightweight agentic tasks. 1M token context window.
Google's frontier model. Leading scores on 13/16 benchmarks. 2M+ context window, native multimodal, Google Search grounding.
IBM
Dense, decoder-only 8B-parameter language model from IBM, part of the Granite 4.1 family. 131K-token context window, designed for enterprise tasks like RAG, summarization, classification, and tool calling. Apache-2.0 open weights.
xAI
xAI's reasoning model with real-time X/Twitter data access and strong analytical capabilities.
xAI
xAI reasoning model accepting text and image inputs with text output. Suited for agentic workflows, instruction-following, and applications requiring high factual accuracy. 1M token context window.
Meta
Meta's open-weight frontier model — free to use, fine-tune, and deploy at any scale.
Meta
Meta's 400B MoE model with 128 experts. Strong multilingual, long-context, and instruction-following capabilities.
Mistral
Mistral
Dense 128B instruction-following model from Mistral AI. Supports text and image inputs with text output. Designed for agentic workflows, coding, and complex reasoning. 256K context window.
NVIDIA
30B-A3B open multimodal model from NVIDIA designed to function as a perception and context sub-agent in enterprise agent systems. Accepts text, image, video, and audio. 256K context. Reasoning-tuned variant.
OpenRouter
High-performance foundation model designed for agentic workloads. Native tool use and long-context support, with strong performance in code generation, automated workflows, and complex instruction execution. 1M+ token context.
Alibaba
Alibaba
Large-scale multimodal language model from Alibaba (April 2026 release). Accepts text, image, and video input with text output. 1M token context window. Tuned for production agentic and reasoning workloads.
Alibaba
Fast, efficient language model from Alibaba's Qwen 3.6 series. Supports text, image, and video input with 1M token context window. Tiered pricing kicks in beyond context thresholds. Optimized for high-throughput, low-latency workloads.
inclusionAI
1T-parameter-scale thinking model with 63B active parameters, built for real-world agent workflows that require strong capability and operational efficiency. Optimized for coding agents and tool use. 256K context window. Available free on OpenRouter.