View overall rankings across various AI models in text-to-text tasks across math, coding, creative writing, and other open-ended domains.
Model performance at each price point
Top performers for their cost
claude-opus-4-7-thinking
Anthropic · Proprietary
1563
$20/M
gpt-5.4-high
OpenAI · Proprietary
1527
$11.88/M
glm-5.1
Z.ai · MIT
1527
$3.65/M
kimi-k2.6
Moonshot · Modified MIT
1519
$3.24/M
mimo-v2.5-pro
Xiaomi · MIT
1517
$2.50/M
gemini-3-flash
Google · Proprietary
1509
$2.38/M
kimi-k2.5-instant
Moonshot · Modified MIT
1506
$1.52/M
deepseek-v4-pro-thinking
DeepSeek · MIT
1500
$0.76/M
gemma-4-31b
Google · Apache 2.0
1498
$0.34/M
deepseek-v4-flash-thinking
DeepSeek · MIT
1486
$0.22/M
gpt-oss-120b
OpenAI · Apache 2.0
1390
$0.14/M
gpt-oss-20b
OpenAI · Apache 2.0
1369
$0.11/M
gemma-3-12b-it
Google · Gemma
1317
$0.11/M
mistral-small-24b-instruct-2501
Mistral · Apache 2.0
1312
$0.07/M
gemma-3-4b-it
Google · Gemma
1274
$0.07/M
llama-3.1-8b-instruct
Meta · Llama 3.1 Community
1259
$0.04/M
llama-3-8b-instruct
Meta · Llama 3 Community
1251
$0.04/M