Reviews and measurements as of 2026-10, 41 models, 1,670 reviews

LLM Leaderboard

Price, latency, and real user reviews, side by side.

Most mentioned, last 7 days

At a glance

Price vs Satisfaction

OpenAIAnthropicGoogleDeepSeekKimi
0%25%50%75%100%$0.1$0.3$1$3$10$30Opus 5.5GPT-6 AstraV4.1 FlashGPT-6.1 SolSonnet 5.5GPT-5.6 Sol
Positive reviewsInput price, $ / 1M tokens
9 questions, 30 secondsNot sure which model fits your product?Find a candidate now and get the details in a report.Find your model in 30 seconds

Head-to-head

Most compared
VS
80%20%
10 comparisons
VS
56%44%
9 comparisons
VS
67%33%
9 comparisons
VS
71%29%
7 comparisons
VS
71%29%
7 comparisons
VS
33%67%
6 comparisons
VS
67%33%
6 comparisons
VS
40%60%
5 comparisons

Model Matrix

Cheapest input first$ / 1M tokens

$ / 1M tokens
OpenAI
Anthropic
Google
DeepSeek
Kimi
$20+3 models
$20+3 models
$103 models
$103 models
$4–$58 models
$4–$58 models
$2–$311 models
$2–$311 models
$1–$24 models
$1–$24 models
$0.5–$15 models
$0.5–$15 models
<$0.57 models
<$0.57 models
Price change / preview / end-of-life notice

AI model comparison FAQ

Which models are in this comparison?

41 models from OpenAI GPT, Anthropic Claude, Google Gemini, DeepSeek and Kimi sit in one table. Open any model to see its price, response time, max input and output tokens, and user reviews together.

How are prices compared?

Prices are API prices in US dollars per 1 million tokens, shown separately for input and output.

How is response time compared?

Response time is the time to receive 500 output tokens, so every model is compared on the same basis.

Can I just pick the top-ranked model?

Rankings and prices are good for narrowing candidates. Whether a model works for your product shows up in how your own users respond. ABTO lets you route a share of users to a candidate model and compare payment, retention and cost. Cutting LLM cost without losing results