20 public AI API SKUs: starting-price study, 06 November 2026

A dated study of 20 public API SKUs, with modality counts and separate price distributions by billing unit.

Published 2026-11-06 · Updated 2026-11-06 · CheapAIAPI

This study describes a dated public pricing snapshot. See models and pricing for current public rates and account access.

Catalog coverage by modality

In CheapAIAPI's 2026-11-06 public starting-price snapshot, the catalog contains 20 unique SKUs: 10 image, 4 text, and 6 video. Each SKU counts once.

ModalityUnique SKUsShare of all unique SKUs
image1050%
text420%
video630%

Image starting prices

In CheapAIAPI's 2026-11-06 public starting-price snapshot, across 10 image SKUs, the lowest starting price is $0.024 and the median is $0.04 per generation. Lowest-price SKUs: nano-banana, nano-banana-2-lite-1k.

UnitSKU countLowestMedianLowest SKU(s)
USD per image generation10$0.024$0.04nano-banana, nano-banana-2-lite-1k

Text input and output prices

In CheapAIAPI's 2026-11-06 public starting-price snapshot, across 4 text SKUs, median input pricing is $0.7875 and median output pricing is $4.4625 per million tokens. Lowest input price is $0.175 (gemini-3.1-flash-lite); lowest output price is $1.05 (gemini-3.1-flash-lite).

UnitSKU countLowestMedianLowest SKU(s)
USD per 1M input tokens4$0.175$0.7875gemini-3.1-flash-lite
USD per 1M output tokens4$1.05$4.4625gemini-3.1-flash-lite

Video resolution-rate distribution

In CheapAIAPI's 2026-11-06 public starting-price snapshot, the 6 video SKUs expose 17 model-resolution rates. The lowest is $0.0427 (wan-3.0 / 480p) and the median is $0.171 per generated output second.

UnitModel-resolution rate countLowestMedianLowest model-resolution pair(s)
USD per output second17$0.0427$0.171wan-3.0 / 480p

Underlying image and text starting prices

Exact source decimal values, USD. Each listed image SKU is one source catalog entry. Text input and output rates are per million tokens.

SKUModalityPricing unitImage priceInput priceOutput price
gemini-3.1-flash-litetextper_million_tokens—0.1751.05
gemini-3.1-protextper_million_tokens—1.48.4
gemini-3.5-flashtextper_million_tokens—1.056.3
gemini-3.7-flashtextper_million_tokens—0.5252.625
gpt-image-2.5imageper_call0.04——
gpt-image-2.6imageper_call0.04——
nano-bananaimageper_call0.024——
nano-banana-2-lite-1kimageper_call0.024——
nano-banana-2.1-1kimageper_call0.028——
nano-banana-2.1-2kimageper_call0.038——
nano-banana-2.1-4kimageper_call0.06——
nano-banana-pro-1kimageper_call0.07——
nano-banana-pro-2kimageper_call0.08——
nano-banana-pro-4kimageper_call0.11——

Underlying video rates

Exact source decimal values, USD per generated output second. The public source formula is resolution_rate_usd * output_duration_seconds. Reference-video billing is outside this snapshot's formula and these findings.

SKUResolutionUSD per output second
seedance-2.01080p0.55
seedance-2.0480p0.115
seedance-2.04k1.25
seedance-2.0720p0.25
seedance-2.0-fast480p0.08
seedance-2.0-fast720p0.18
seedance-2.0-mini480p0.054
seedance-2.0-mini720p0.115
seedance-2.51080p0.93
seedance-2.5480p0.175
seedance-2.5720p0.375
wan-3.01080p0.171
wan-3.0480p0.0427
wan-3.0720p0.0855
wan-3.0-prime1080p0.2674
wan-3.0-prime480p0.0669
wan-3.0-prime720p0.1337

Methodology and limitations

Observed https://cheapaiapi.org/pricing.json at 2026-11-06T15:39:29.688917+00:00. Source status: database_backed_snapshot; prices_may_be_stale=false. Scope is public new-customer starting prices, prepaid usage-based, USD. Identical SKU duplicates count once; conflicting duplicates and invalid prices are rejected. Each SKU has equal weight in its modality. Video distribution counts each model-resolution pair once. Medians average the middle pair for an even sample. Catalog-share percentages use all unique SKUs as denominator and round to two decimals. At least 20 distinct SKUs are required for publication. Each price distribution requires two observations. Rates from different billing units are never pooled. These descriptive statistics describe public starting-price entries. Quality, latency, popularity, competitor savings, and account-specific prices are outside the study. Consult current pricing.json and authenticated GET /v1/models for current account terms.

Frequently asked questions

How many public API SKUs are in this snapshot?

In CheapAIAPI's 2026-11-06 public starting-price snapshot, the catalog contains 20 unique SKUs: 10 image, 4 text, and 6 video. Each SKU counts once.

What are the lowest and median image starting prices?

In CheapAIAPI's 2026-11-06 public starting-price snapshot, across 10 image SKUs, the lowest starting price is $0.024 and the median is $0.04 per generation. Lowest-price SKUs: nano-banana, nano-banana-2-lite-1k.

What are the median input and output token prices?

In CheapAIAPI's 2026-11-06 public starting-price snapshot, across 4 text SKUs, median input pricing is $0.7875 and median output pricing is $4.4625 per million tokens. Lowest input price is $0.175 (gemini-3.1-flash-lite); lowest output price is $1.05 (gemini-3.1-flash-lite).

How many video rates are listed and what is their median?

In CheapAIAPI's 2026-11-06 public starting-price snapshot, the 6 video SKUs expose 17 model-resolution rates. The lowest is $0.0427 (wan-3.0 / 480p) and the median is $0.171 per generated output second.

Suggested citation

CheapAIAPI. (2026-11-06). 20 public AI API SKUs: starting-price study, 06 November 2026. https://cheapaiapi.org/research/api-pricing.

https://cheapaiapi.org/research/api-pricing

Sources

All research · Read as Markdown