Skip to content

GPT Image 2 API

−41% vs list rate

OpenAI's image model with best-in-class prompt following

gpt-image-2Commercial useSynchronousImage inputText renderingimageLive price on TokPort: $0.10 per image

Price on TokPort

$0.10 / image

Indicative first-party list rate: $0.17 / image. Live rates: pricing.

Monthly at scale

1,000 images$100
10,000 images$1,000
Calculate your exact spend →

About GPT Image 2

GPT Image 2 is OpenAI's image generation model, delivering exceptional prompt adherence, accurate in-image text and native editing via image reference. Ideal when the image must match a complex instruction exactly — diagrams, UI mockups, posters and infographics.

  • Best-in-class prompt adherence
  • Accurate in-image typography
  • Image reference / editing support
  • Same OpenAI images API you already know

Specifications

Max resolution
2048 × 2048
Typical latency
50–90 s
Billing unit
per generated image
Input
text · image reference
Model id
gpt-image-2

One synchronous request against the standard OpenAI images endpoint:

curl https://api.tokport.io/v1/images/generations \
  -H "Authorization: Bearer $TOKPORT_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"gpt-image-2","prompt":"a neon koi swimming through clouds, cinematic lighting","size":"1024x1024","n":1}'

GPT Image 2 FAQ

How much does GPT Image 2 cost per image?

TokPort charges $0.10 per image for GPT Image 2. The indicative first-party API list rate is about $0.17 per image — roughly 41% less. Check the live pricing page for current rates.

Is GPT Image 2 available via an OpenAI-compatible API?

Yes. Generate images with a standard POST /v1/images/generations request and model "gpt-image-2". The official OpenAI SDKs work unchanged — just point them at our base URL.

What is GPT Image 2 best at?

GPT Image 2 is OpenAI's image generation model, delivering exceptional prompt adherence, accurate in-image text and native editing via image reference. Ideal when the image must match a complex instruction exactly — diagrams, UI mockups, posters and infographics. It is strongest for prompt adherence, text in image, instruction edits.

Made with GPT Image 2

All prompts →
Create a 9:16 editorial poster in the HUMAN FRAGMENTS visual system. Use oversiz

Create a 9:16 editorial poster in the HUMAN FRAGMENTS visual system. Use oversiz

GPT Image 2
Anime hiker scrambling up a ridge at sunrise

Anime hiker scrambling up a ridge at sunrise

GPT Image 2 · TokPort
Flash-lit iPhone close-up portrait

Flash-lit iPhone close-up portrait

GPT Image 2 · TokPort
Cel-shaded mountain climb above a lakeside town

Cel-shaded mountain climb above a lakeside town

GPT Image 2 · TokPort
Photorealisticlive actionscifi filmstll. The mecha must look EXACTLY

Photorealisticlive actionscifi filmstll. The mecha must look EXACTLY

GPT Image 2
Build an intimate fashion magazine cover with the warmth of a personal snapshot

Build an intimate fashion magazine cover with the warmth of a personal snapshot

GPT Image 2
Cartoon cat from a reference photo

Cartoon cat from a reference photo

GPT Image 2 · TokPort

Related models