Live survey286 models · 13 providers · 19,992 auditsAudited verification-before-completion · SAFE →What changedHow we verifyllms.txt

669 head-to-head comparisons · 66 provider comparisons

AI model comparison

Compare AI models on price per 1M tokens, context window, input types and real monthly cost. Pick a pair below, or open any model page and follow its “Compare” links. Every number links to its source.

Ask an AI about this pageGChatGPTCClaudePPerplexityGeGeminiXGrok
More
CoCopilotAIGoogle AI Mode
Cite

Cite this page

ModelsAtlas. “AI model comparison”. Retrieved 12 Oct 2026. https://modelsatlas.com/compare

Head-to-head comparisons

Rival models in the same atlas square, dearest first

Claude Opus 4.1 vs o1$15.00 · $15.00Claude Opus 4 vs o1$15.00 · $15.00Claude Opus 4.6 vs Qwen3.8 Max Prime$5.00 · $4.00Claude Opus 4.7 vs Qwen3.8 Max Prime$5.00 · $4.00Claude Opus 4.8 vs Qwen3.8 Max Prime$5.00 · $4.00Claude Opus 5 vs Qwen3.8 Max Prime$5.00 · $4.00Claude Opus 5.5 vs Qwen3.8 Max Prime$4.00 · $4.00Claude Sonnet 4.5 vs Qwen3.8 Max Prime$3.00 · $4.00Claude Sonnet 4.6 vs Qwen3.8 Max Prime$3.00 · $4.00Claude Haiku 4.5 vs Command A$1.00 · $2.50Command A vs Nano Banana Pro (Gemini 3 Pro Image)$2.50 · $2.00Command A vs o3$2.50 · $2.00

Prices: input per 1M tokens, standard tier.

Compare within a maker

OpenAI · 53 pairs

Qwen (Alibaba) · 45 pairs

Anthropic · 34 pairs

Google · 17 pairs

Compare providers

66 pairs

Model makers

All 66 provider comparisons

Compare by price

Cheapest modelsFree modelsBatch pricingPricing by provider

How comparisons work

Where do the prices come from?
From vendor pricing pages and our own catalogue, each stored with the date we retrieved it; the newest on this site is 12 Oct 2026. Monthly-cost figures are computed from those prices and marked as derived.
Why is there one page per pair?
Each pair has a single address with the two models in alphabetical order, so “A vs B” and “B vs A” lead to the same page.
Which pairs get a page?
Each model is compared with the models nearest it on the atlas: the same grid square first, then neighbouring squares by price. A pair of a free small model and a frontier model is a page nobody needs.
Do you run benchmarks?
No. We compare recorded facts — price, context window, input types, tiers. Where we give an opinion it is labelled Editorial.

Keep going in your AI

Opens with this page as the source. We add nothing else to the prompt.