GPT Image 2 or 2.5 on Sume: 27 cents vs 7 cents per high image
At the catalog high/1024 rate, GPT Image 2 is $0.211 list and 27 cents billable on Sume; GPT Image 2.5 is $0.0527 list and 7 cents. 100 images: $27 vs $7.
The short answer
On Sume, a high-quality 1024x1024 image costs 7 cents on GPT Image 2.5 and 27 cents on GPT Image 2. For 100 images that is $7.00 against $27.00, a difference of $20.00. Both prices are the catalog's high/1024 display rate times 1.25, rounded up to the cent.
GPT Image 2.5 also has cheaper tiers: 1 cent at low and 2 cents at medium for 1024-class images, so it can serve both drafts and finals.
The two prices side by side
The catalog note for both models says the admission estimate uses OpenAI's image-token estimate by quality and size, and that the per-image list shown is the high/1024 display rate. Compare like for like: high against high.
| Model | Catalog list | x 1.25 | Billed per image | 100 images | 1,000 images |
|---|---|---|---|---|---|
| GPT Image 2 | $0.211 | $0.26375 | $0.27 | $27.00 | $270.00 |
| GPT Image 2.5 | $0.0527 | $0.065875 | $0.07 | $7.00 | $70.00 |
Why the difference is so large
The Sume images page documents the 2.5 rate card: $30 per million output image tokens, $8 per million input image tokens and $5 per million input text tokens, with output estimated from OpenAI's size and quality calculator. The 2.5 high tier at 1024x1024 comes to $0.0527 in the catalog. The GPT Image 2 catalog row shows $0.211 at the same high/1024 point, which is 4.0x the 2.5 list price.
The arithmetic: $0.211 x 1.25 = $0.26375, rounded up to $0.27. And $0.0527 x 1.25 = $0.065875, rounded up to $0.07. Because both are rounded up per image, 100 images are exactly 100 times the per-image price.
What the gap costs at working volumes
The difference per image is 27 - 7 = 20 cents. At 1,000 images a month that is $200 against $70, a $130 gap. At 10,000 images it is $2,700 against $700. At 50 images, the size of a campaign test, it is $13.50 against $3.50, which is a gap of $10.
These figures hold at high quality and 1024x1024 only. Other sizes and quality tiers re-price through the token estimate, so use them as a ceiling for a high-quality plan rather than a quote for every request.
When 2 is still the right call
Price is not the only axis. The images page says you can still select ChatGPT Image 2. If your pipeline pins GPT Image 2 for a look you have already approved, that is a reason to keep it; if you have not tested 2.5 on the same prompts, a 20-image side-by-side costs 20 x 7 = $1.40 at high on 2.5.
Check the live numbers with GET /v1/images/models; the Images page explains the pricing lines.
Sources
Related posts
More in Pricing
- Grok Imagine video 1.5: 1.25 cents a second at any resolution
grok-imagine-video-1.5 bills $0.0125 a second on Sume at 480p or 720p. A table of 4 to 15 second clips and what 100 and 1,000 of them cost.
- H3 Max draft ladder: 15 shots, 5 reruns, 2 finals for $14.70 on Sume
Price a 480p to 768p to 1080p ladder of eight-second H3 Max shots: $14.70 on Sume versus $24.00 all at 1080p, with fal's 40% promo ending October 15 shown.
- H3 Max Recast: what a 30-second people swap costs at 768p and 1080p
H3 Max Recast on Sume swaps the people in a 5 to 30 s source video for 1 to 4 photos. 30 s costs $11.25 at 768p and $16.875 at 1080p.
- H3 Max Recast: a 30-second, 4-person swap costs $11.25 at 768p
Sume bills a 30-second H3 Max Recast at 768p for $11.25 and 1080p for $16.88. Reference photos add nothing; fal's page says the same. Arithmetic inside.
Written by Sume