AI model comparison: prices, quality and speed of every major model

Model Finder

Quadrant charts: what you get for the price

Generally available text models with published figures. Dashed lines mark the median, so the top left of each chart is above-median results for below-median price. Hover or tap a dot for its source.

AI API pricing comparison: official prices per 1M tokens

26 generally available text models with an official token price and an independent quality score, cheapest first. Prices are each lab's standard list price in US dollars, checked 1 Nov 2026. Blended price weights three input tokens to one output token.

Quality index: Artificial Analysis Intelligence Index at each model's best tested setting. Long-context, batch and cached rates differ, see each model page.

ChatGPT vs Claude vs Gemini: the models compared

ChatGPT, Claude and Gemini are apps. Each runs on a family of models from its lab, OpenAI, Anthropic and Google. Below are the highest-scoring and the cheapest scored model from each, with official API prices.

Index: Artificial Analysis Intelligence Index. App plan prices and limits are on each maker's pricing page. Chat assistants and their agents are compared in the agent finder.

Best AI models by independent quality score

The 10 highest-scoring text models on the Artificial Analysis Intelligence Index, with their list price. This is one independent measure of general capability, not a verdict for every job. For coding, compare the coding agents that use these models.

1. Claude Opus 5.5 (Anthropic): index 58, US$4 in and US$20 out per 1M tokens. Sources: artificialanalysis.ai (opens in a new tab), docs.anthropic.com (opens in a new tab)
2. GPT-6 Astra (OpenAI): index 57, US$10 in and US$50 out per 1M tokens. Sources: artificialanalysis.ai (opens in a new tab), developers.openai.com (opens in a new tab)
3. GPT-6.1 Sol (OpenAI): index 55, US$2 in and US$10 out per 1M tokens. Sources: artificialanalysis.ai (opens in a new tab), developers.openai.com (opens in a new tab)
4. Claude Sonnet 5.5 (Anthropic): index 54, US$2 in and US$10 out per 1M tokens. Sources: artificialanalysis.ai (opens in a new tab), docs.anthropic.com (opens in a new tab)
5. GPT-6.1 Terra (OpenAI): index 51, US$1 in and US$5 out per 1M tokens. Sources: artificialanalysis.ai (opens in a new tab), developers.openai.com (opens in a new tab)
6. Claude Fable 5.1 (Anthropic): index 50, US$10 in and US$50 out per 1M tokens. Sources: artificialanalysis.ai (opens in a new tab), docs.anthropic.com (opens in a new tab)
7. Gemini 4 Argon (Google): index 50, US$2 in and US$10 out per 1M tokens. Sources: artificialanalysis.ai (opens in a new tab), ai.google.dev (opens in a new tab)
8. Kimi K3.1 Pro (Moonshot): index 46, US$2 in and US$10 out per 1M tokens. Sources: artificialanalysis.ai (opens in a new tab), platform.kimi.ai (opens in a new tab)
9. Muse Wonder (Meta): index 43, US$4 in and US$20 out per 1M tokens. Sources: artificialanalysis.ai (opens in a new tab), dev.meta.ai (opens in a new tab)
10. Claude Haiku 5.5 (Anthropic): index 36, US$0.1 in and US$0.5 out per 1M tokens. Sources: artificialanalysis.ai (opens in a new tab), docs.anthropic.com (opens in a new tab)

At a glance, by category

• Text and reasoning

  75 models from 17 labs, 27 with open weights

  Lowest blended list price

  Ministral 3 3B, US$0.10 per 1M docs.mistral.ai (opens in a new tab)

  Highest Artificial Analysis index

  Claude Opus 5.5, 58 artificialanalysis.ai (opens in a new tab)

  Fastest measured output

  Nemotron 3.5 Super VL, 474 tokens/s artificialanalysis.ai (opens in a new tab)

  Largest context

  GPT-6 Astra, 1.05M developers.openai.com (opens in a new tab)

• Image

  9 models from 4 labs, 0 with open weights

  Lowest blended list price

  Nano Banana 2.1 (Gemini image), US$8.63 per 1M ai.google.dev (opens in a new tab)

• Video

  2 models from 2 labs, 0 with open weights

  Lowest blended list price

  Gemini Omni 1.1 Flash (video), US$5.50 per 1M ai.google.dev (opens in a new tab)

• Voice

  17 models from 7 labs, 0 with open weights

  Lowest blended list price

  Gemini 3.8 Flash TTS, US$2.63 per 1M ai.google.dev (opens in a new tab)

• Embeddings

  3 models from 2 labs, 0 with open weights

Facts computed from the table, not a ranking. Limited-access models are left out. Prices for image, video and voice models are per token of that medium, so compare them only within a category.

Explainers

Cost calculator, every priced model

Enter a monthly token volume to see the list-price cost on each model with a published token price.

Ecosystems: who makes these models

Benchmark claims published by the labs

Different tests and settings, so rows are not comparable with each other.

Code examples from official docs

from agents import Agent, Runner

agent = Agent(name="Assistant", instructions="You are a helpful assistant")

result = Runner.run_sync(agent, "Write a haiku about recursion in programming.")
print(result.final_output)

from google import genai

client = genai.Client()

interaction = client.interactions.create(
    model="gemini-4-argon",
    input="Write a three.js script that renders a realistic 3D black hole with gravitational lensing."
)

print(interaction.output_text)

from google.adk import Agent
from google.adk.tools import google_search

agent = Agent(
    name="researcher",
    model="gemini-flash-latest",
    instruction="You help users research topics thoroughly.",
    tools=[google_search],
)

{
  "model": "glm-5.3",
  "thinking": { "type": "enabled" },
  "reasoning_effort": "max"
}

Watch

Official uploads. Nothing loads from YouTube until you press play.

▶

▶