Protegee
Log in Sign up

Model guide

AI model pricing, at a glance

Representative API list prices for the leading AI video and image models in late 2026, drawn as a cost chart. The bar shows the relative cost, the number shows the exact rate.

By the Protegee team · October 21, 2026 · Representative list prices

This page is the data. For the discussion behind the numbers, how per-second, token, and credit billing differ, what drives the rate, and why cost per finished post beats the sticker price, read AI video pricing and AI image pricing. Each bar is scaled to the most expensive rate within its own table, because video is billed per second and images per image, which are different units and cannot share one scale.

AI video model pricing, per second

For roughly 720p image-to-video, the format most social clips start from. Bar length is scaled to the top of this table, about $0.40 per second. Read the full context in AI video pricing.

ModelProviderPrice per secondNotes
Veo 3.1 Google $0.05 to $0.40 Lite 720p about $0.05, Fast about $0.10, Standard $0.40 with generated audio
Kling 2.5 Turbo Pro Kuaishou About $0.07 $0.35 per 5-second clip plus about $0.07 per extra second, silent, no audio
Seedance 1 Pro ByteDance About $0.05 to $0.12 Token-billed at $2.50 per 1M video tokens, roughly $0.05 per second at 720p rising to $0.12 at 1080p

Representative API list prices, as of October 2026. These are ballpark list rates, not quotes. Prices move often and vary by resolution, tier, clip length, and whether audio is generated, so check the provider's own page before you plan around a number.

Omitted
  • Sora 2 (OpenAI), because OpenAI retired its video API on September 24, 2026.
  • Runway Gen-4 Turbo and Luma Ray 2, because both price in credits with no clean published per-second dollar rate, so any figure would be an estimate.

Sources, read October 2026: Veo 3.1 from Google Gemini API pricing; Kling 2.5 Turbo Pro from the fal.ai model page; Seedance 1 Pro from the fal.ai model page.

AI image model pricing, per image

For a 1024px image, the standard size for most social and product work. Bar length is scaled to the top of this table, about $0.167 per image. Read the full context in AI image pricing.

ModelProviderPrice per imageNotes
GPT Image 1 Mini OpenAI $0.005 to $0.036 Low $0.005, medium $0.011, high $0.036
GPT Image 1 OpenAI $0.011 to $0.167 Low $0.011, medium $0.042, high $0.167
GPT Image 2 OpenAI Token-billed No public per-image rate OpenAI publishes token rates only for this model, not a per-image table. Cost depends on quality and output size, so there is no single list figure
Gemini 3.1 Flash Image Google $0.067 Also known as Nano Banana 2, flat per-image rate with no quality dial
Gemini 3 Pro Image Google $0.134 Also known as Nano Banana Pro, flat per-image rate with output up to 2K

Representative API list prices, as of October 2026. These are ballpark list rates, not quotes. Prices move often and vary by quality tier and output size, so check the provider's own page before you plan around a number.

Omitted
  • Gemini 2.5 Flash Image, the old Nano Banana, because it was deprecated and shut down on October 2, 2026, replaced by the Gemini 3 line above.
  • Midjourney, FLUX, Ideogram, Recraft, and Firefly, because their plans are subscription or credit based rather than a clean published per-image dollar rate.

Sources, read October 2026: GPT Image 1 and GPT Image 1 Mini from the OpenAI image generation guide; the Gemini image models from Google Gemini API pricing. GPT Image 2 is billed by token with no published per-image table, so it is shown without an invented figure. Prices are representative list rates and move often, so confirm with the provider before you budget.

The right model, picked for you

Protegee renders through the leading AI video and image models and picks the one that fits each brief, so you are never locked to a single engine or its price. It makes the content, attaches your brand, and posts to every account on a schedule. You never choose a model or watch a meter.

Sign up