Image prices, on their real billing terms.
Compare offers billed per finished image. Read token-metered models separately: their rate is not the cost of one generated image.
Read the unit first.Rates for different resolution, quality and input modes are not equivalent. A Runway Dev route is billed by Runway Dev, even when another lab created the model.
01 / FIXED OUTPUT UNIT
Priced per generated image
Published rates, lowest first. Modes and quality differ.
| Model | USD / image | Output mode | Billing seller | Source and date |
|---|---|---|---|---|
| Muse Imagemuse_image | $0.01 / image | Any resolution | Runway DevModel by Meta | Official rateVerified |
| Gen-4 Image Turbogen4_image_turbo | $0.02 / image | Any resolution | Runway DevModel by Runway | Official rateVerified |
| Seedream 5.0 Flashseedream5_flash | $0.02 / image | Any resolution | Runway DevModel by ByteDance | Official rateVerified |
| Stable Image Corev2beta/stable-image/generate/core | $0.03 / image | Generate API; standard request. Resolution and output limits follow the linked endpoint. | Stability AI | Official rateVerified |
| Stable Diffusion 3.5 Mediumsd3.5-medium | $0.035 / image | Generate API; standard request. Resolution and output limits follow the linked endpoint. | Stability AI | Official rateVerified |
| Nano Banana 2 Litegemini_image3.1_flash_lite | $0.04 / image | 1K output | Runway DevModel by Google | Official rateVerified |
| Seedream 5.0 Liteseedream5_lite | $0.04 / image | Any resolution | Runway DevModel by ByteDance | Official rateVerified |
| Stable Diffusion 3.5 Large Turbosd3.5-large-turbo | $0.04 / image | Generate API; standard request. Resolution and output limits follow the linked endpoint. | Stability AI | Official rateVerified |
| Stable Diffusion 3.5 Largesd3.5-large | $0.065 / image | Generate API; standard request. Resolution and output limits follow the linked endpoint. | Stability AI | Official rateVerified |
| Stable Image Ultrav2beta/stable-image/generate/ultra | $0.08 / image | Generate API; standard request. Resolution and output limits follow the linked endpoint. | Stability AI | Official rateVerified |
The table orders published rates, not model quality or equivalent output. Check the named mode before comparing.
02 / TOKEN METER
Priced per million tokens
These figures are not prices per generated image.
Scroll to read the text-token rates and source.
| Model | USD / 1M image tokens | USD / 1M text tokens | Source and date | ||||
|---|---|---|---|---|---|---|---|
| Input | Output | Cached input | Input | Output | Cached input | ||
| gpt-image-2.5-sunburstOpenAI | $8.00 | $30.00 | $2.00 | $5.00 | — | $1.25 | Official rateVerified |
| gpt-image-2.5-flareOpenAI | $8.00 | $30.00 | $2.00 | $5.00 | — | $1.25 | Official rateVerified |
Total request cost needs the actual tokens used in each billed dimension. A dash means the seller did not publish that rate; it does not mean free. Cache rates may require a specific API mode.