ModelServerModelServer
Rankings

The best AI models,
ranked.

Pick a category to see how the top AI models stack up — by price, speed, context, or capability.

Cheapest input pricing

Lowest cost per million input tokens.

  1. 1
    GPT-4o mini
    OpenAI
    $0.15 / 1M tok
  2. 2
    DeepSeek V3
    DeepSeek
    $0.27 / 1M tok
  3. 3
    DeepSeek R1
    DeepSeek
    $0.55 / 1M tok
  4. 4
    Claude Haiku 4.5
    Anthropic
    $1.00 / 1M tok
  5. 5
    Gemini 2.5 Pro
    Google
    $1.25 / 1M tok
  6. 6
    GPT-4o
    OpenAI
    $2.50 / 1M tok

Cheapest output pricing

Lowest cost per million output tokens.

  1. 1
    GPT-4o mini
    OpenAI
    $0.60 / 1M tok
  2. 2
    DeepSeek V3
    DeepSeek
    $1.10 / 1M tok
  3. 3
    DeepSeek R1
    DeepSeek
    $2.19 / 1M tok
  4. 4
    Claude Haiku 4.5
    Anthropic
    $5.00 / 1M tok
  5. 5
    GPT-4o
    OpenAI
    $10.00 / 1M tok
  6. 6
    Gemini 2.5 Pro
    Google
    $10.00 / 1M tok

Largest context window

Models that handle the longest inputs in a single call.

  1. 1
    Gemini 2.5 Pro
    Google
    1M
  2. 2
    Claude Opus 4
    Anthropic
    200K
  3. 3
    Claude Sonnet 4
    Anthropic
    200K
  4. 4
    Claude Haiku 4.5
    Anthropic
    200K
  5. 5
    o3
    OpenAI
    200K
  6. 6
    GPT-4o
    OpenAI
    128K

Top reasoning models

Best for math, code, and step-by-step problem solving.

  1. 1
    Claude Opus 4
    Anthropic
    flagship
  2. 2
    GPT-4o
    OpenAI
    flagship
  3. 3
    o3
    OpenAI
    reasoning
  4. 4
    Gemini 2.5 Pro
    Google
    flagship
  5. 5
    DeepSeek R1
    DeepSeek
    reasoning
AI Model Rankings — By price, speed and capability — ModelServer