Pick the right AI model
in 60 seconds.
Independent, opinionated rankings of every major LLM. Compare price, context, and real-world fit for coding, writing, agents, and research — without wading through a dozen vendor sites.
Most-compared this month
Top models for coding
Claude Opus 4.7
Anthropic's flagship. Deep reasoning, long-horizon agents.
DeepSeek V3
Open-weights reasoning at bargain pricing.
Claude Sonnet 4.6
Sonnet strikes back. The price-performance sweet spot.
Top models for bulk workloads
Top models for reasoning
All tracked models
Claude Opus 4.7
Anthropic's flagship. Deep reasoning, long-horizon agents.
Claude Sonnet 4.6
Sonnet strikes back. The price-performance sweet spot.
Claude Haiku 4.5
Fast, cheap, surprisingly smart.
GPT-5
OpenAI's flagship. Omni-modal, omni-purpose.
GPT-5 Mini
GPT-5 shrunk for scale.
GPT-4.1
Long-context workhorse still on the menu.
o4
Reasoning model. Thinks before it speaks.
Gemini 3 Pro
Google's flagship. Massive context, native multimodality.
Gemini 3 Flash
Cheap, fast, 2M-token context.
Grok 4
Real-time-aware, unfiltered, strong reasoner.
Llama 4 405B
Open-weights heavyweight. Self-host or run anywhere.
Llama 4 70B
Mid-size open model. Excellent price-performance when self-hosted.
DeepSeek V3
Open-weights reasoning at bargain pricing.
Mistral Large 3
European flagship with strong multilingual chops.
Mistral Small 3
Small, open-weights, surprisingly sharp.
Qwen 3 Max
Alibaba's flagship. Top-tier multilingual and coding.
Qwen 3 Coder
Open-weights coding specialist.
Command R+
RAG-tuned enterprise workhorse.
GPT-OSS 20B
OpenAI's open-weights return.
Gemini Nano 2
On-device first. Free inference, private by default.