Last checked: 2026-09-28

OpenAI: API Prices and Provider Profile

Type
Official Official: the model's own developer
Full name
OpenAI API
Website
https://developers.openai.com/api
Pricing page
developers.openai.com pricing
Models priced
4 (11 price rows)

Models listed: GPT Image 1 Mini, GPT Image 2, GPT Image 2.5 Flare, GPT Image 2.5 Sunburst.

Published prices (11)

Published prices — not yet independently tested. Figures are copied from each provider's public pricing page or pricing API on 2026-09-28. We have not run test generations yet, so a listed price says nothing about reliability or output quality. Every price links to its source.

Image models (11)

ModelTypeSpecPublished price$ / image× officialConfidenceSourceNote
GPT Image 1 MiniOfficialLow · 1024×1024Listed as: GPT Image 1 Mini$0.005 / image$0.005001.00×Pricing pagedevelopers.openai.com
Note
Per-image price stated on model page; 1024x1536 / 1536x1024 = $0.006. Token rates: image output $8/1M, image input $2.50/1M, text input $2/1M. Batch API: 50% off (image output $4/1M). Input tokens billed extra.
GPT Image 1 MiniOfficialMedium · 1024×1024Listed as: GPT Image 1 Mini$0.011 / image$0.01101.00×Pricing pagedevelopers.openai.com
Note
Per-image price stated on model page; 1024x1536 / 1536x1024 = $0.015. Token rates: image output $8/1M, image input $2.50/1M, text input $2/1M. Batch API: 50% off (image output $4/1M). Input tokens billed extra.
GPT Image 1 MiniOfficialHigh · 1024×1024Listed as: GPT Image 1 Mini$0.036 / image$0.03601.00×Pricing pagedevelopers.openai.com
Note
Per-image price stated on model page; 1024x1536 / 1536x1024 = $0.052. Token rates: image output $8/1M, image input $2.50/1M, text input $2/1M. Batch API: 50% off (image output $4/1M). Input tokens billed extra.
GPT Image 2OfficialLow · 1024×1024Listed as: GPT Image 2$0.006 / image$0.006001.00×Pricing pagedevelopers.openai.com
Note
Per-image output cost stated in OpenAI's 'Calculating costs' table; 1024x1536 / 1536x1024 = $0.005. Token rates: image output $30/1M, image input $8/1M, text input $5/1M. Batch API: 50% off token rates (image output $15/1M). Supports sizes up to 3840px edge (2K/4K). Input tokens billed extra.
GPT Image 2OfficialMedium · 1024×1024Listed as: GPT Image 2$0.053 / image$0.05301.00×Pricing pagedevelopers.openai.com
Note
Per-image output cost stated in OpenAI's 'Calculating costs' table; 1024x1536 / 1536x1024 = $0.041. Token rates: image output $30/1M, image input $8/1M, text input $5/1M. Batch API: 50% off token rates (image output $15/1M). Supports sizes up to 3840px edge (2K/4K). Input tokens billed extra.
GPT Image 2OfficialHigh · 1024×1024Listed as: GPT Image 2$0.211 / image$0.2111.00×Pricing pagedevelopers.openai.com
Note
Per-image output cost stated in OpenAI's 'Calculating costs' table; 1024x1536 / 1536x1024 = $0.165. Token rates: image output $30/1M, image input $8/1M, text input $5/1M. Batch API: 50% off token rates (image output $15/1M). Supports sizes up to 3840px edge (2K/4K). Input tokens billed extra.
GPT Image 2.5 FlareOfficialMedium · 1024×1024Listed as: GPT Image 2.5 Flare$0.01317 / image$0.013171.00×Pricing pagedevelopers.openai.com
Note
Token-billed: image output $30/1M tokens (image input $8/1M, text input $5/1M, cached $2/$1.25). Per-image figure = output-token cost only, computed with OpenAI's own GPT Image 2.5 calculator formula on the image-generation guide (439 output tokens at 1024x1024 medium); excludes prompt/input tokens. Calculator treats Sunburst and Flare identically. Batch API not supported for this model. Qualities: low/medium/high/xhigh/max. Flare = 'fastest' GPT Image 2.5 variant; same token rates as Sunburst, so low/high/max equal the Sunburst rows per the calculator.
GPT Image 2.5 SunburstOfficialLow · 1024×1024Listed as: GPT Image 2.5 Sunburst$0.00588 / image$0.005881.00×Pricing pagedevelopers.openai.com
Note
Token-billed: image output $30/1M tokens (image input $8/1M, text input $5/1M, cached $2/$1.25). Per-image figure = output-token cost only, computed with OpenAI's own GPT Image 2.5 calculator formula on the image-generation guide (196 output tokens at 1024x1024 low); excludes prompt/input tokens. Calculator treats Sunburst and Flare identically. Batch API not supported for this model. Qualities: low/medium/high/xhigh/max.
GPT Image 2.5 SunburstOfficialMedium · 1024×1024Listed as: GPT Image 2.5 Sunburst$0.01317 / image$0.013171.00×Pricing pagedevelopers.openai.com
Note
Token-billed: image output $30/1M tokens (image input $8/1M, text input $5/1M, cached $2/$1.25). Per-image figure = output-token cost only, computed with OpenAI's own GPT Image 2.5 calculator formula on the image-generation guide (439 output tokens at 1024x1024 medium); excludes prompt/input tokens. Calculator treats Sunburst and Flare identically. Batch API not supported for this model. Qualities: low/medium/high/xhigh/max.
GPT Image 2.5 SunburstOfficialHigh · 1024×1024Listed as: GPT Image 2.5 Sunburst$0.05268 / image$0.052681.00×Pricing pagedevelopers.openai.com
Note
Token-billed: image output $30/1M tokens (image input $8/1M, text input $5/1M, cached $2/$1.25). Per-image figure = output-token cost only, computed with OpenAI's own GPT Image 2.5 calculator formula on the image-generation guide (1756 output tokens at 1024x1024 high); excludes prompt/input tokens. Calculator treats Sunburst and Flare identically. Batch API not supported for this model. Qualities: low/medium/high/xhigh/max.
GPT Image 2.5 SunburstOfficialMax · 1024×1024Listed as: GPT Image 2.5 Sunburst$0.2107 / image$0.21071.00×Pricing pagedevelopers.openai.com
Note
Token-billed: image output $30/1M tokens (image input $8/1M, text input $5/1M, cached $2/$1.25). Per-image figure = output-token cost only, computed with OpenAI's own GPT Image 2.5 calculator formula on the image-generation guide (7024 output tokens at 1024x1024 max); excludes prompt/input tokens. Calculator treats Sunburst and Flare identically. Batch API not supported for this model. Qualities: low/medium/high/xhigh/max. xhigh = 3122 tokens = $0.09366.

× official = normalized price ÷ the cheapest official price for the same model and spec. ~ marks a relaxed match: tier, resolution or audio is unspecified on one side, and every compatible official spec costs the same. — = no unambiguous official price to compare.

Confidence: “Pricing page” = read on the provider’s own pricing or docs page; “Pricing API” = read from a public API response. “BYO” tiers are bring-your-own-account estimates, not per-call prices.

Affiliate program

Program
No affiliate or referral program found

Checked 2026-09-28. This site contains no affiliate links; affiliate programs never affect our data or rankings.

Trust signals

Trustpilot
Not checked

Reported issues and terms

No issues recorded in our 2026-09-28 check.