Simple, transparent pricing.
Pay only for what you use.
API Pricing
Search and filter our model catalog. Powered by successful output only.
Values in credit: 1 credit = $1.00 · closed-source model rates in USD
Nunchux Optimized Models
Per megapixel of output image
| Model | Category | Radical Speed | Radical Value |
|---|---|---|---|
| FLUX.2 Klein 4B | T2I | 0.003 | 0.0006 |
| FLUX.2 Klein 9B | T2I | 0.009 | 0.006 |
| FLUX.1 Schnell | T2I | 0.0025 | 0.0006 |
| FLUX.1 dev | T2I | 0.01 | |
| FLUX.2 dev | T2I | 0.007 | |
| Qwen Image | T2I | 0.004 | |
| Qwen Image 2.1 | T2I | 0.01 | |
| Z-Image Turbo | T2I | 0.0015 | 0.0012 |
| Krea 2 Turbo | T2I | 0.007 | 0.005 |
| FLUX.2 Klein 4B Edit | I2I | 0.004 | 0.0008 |
| FLUX.2 Klein 9B Edit | I2I | 0.01 | 0.007 |
| Qwen Image Edit | I2I | 0.02 | |
| Qwen Image 2512 Lightning | T2I | 0.01 | 0.006 |
| Qwen Image Edit 2511 Lightning | I2I | 0.015 | 0.012 |
| Qwen Image 2.1 Edit | I2I | 0.012 | |
| HiDream O1 | T2I | 0.003 | 0.002 |
| HiDream O1 Flash | T2I | 0.0015 | 0.001 |
Per image
| Model | Category | Turbo | Balanced | Quality |
|---|---|---|---|---|
| Ideogram 4 | T2I | 0.03 | 0.06 | 0.1 |
| Ideogram 4 Flash | T2I | 0.03 | 0.06 | 0.1 |
Per second of output video
| Model | Category | 720P | 1080P |
|---|---|---|---|
| LTX-2.3 (Radical Speed) | T2V / I2V | 0.015 | 0.03 |
| LTX-2.3 (Radical Value) | T2V / I2V | 0.012 | 0.0225 |
| LTX-2.5 (Radical Speed) | T2V / I2V | 0.02 | 0.04 |
| LTX-2.5 (Radical Value) | T2V / I2V | 0.015 | 0.03 |
Per second of output video
(Reference images: 5 free per generation, then per image: MiniMax H3 Fast (Radical Speed) 0.011 (480P) / 0.015 (768P), MiniMax H3 Fast (Radical Value) 0.009 (480P) / 0.012 (768P), MiniMax H3 Fast (Radical Speed) 0.011 (480P) / 0.015 (768P), MiniMax H3 Fast (Radical Value) 0.009 (480P) / 0.012 (768P). Reference audio is free)
| Model | Category | Radical Speed | Radical Value |
|---|---|---|---|
| MiniMax H3 Fast (480P) | T2V | 0.022 | 0.018 |
| MiniMax H3 Fast (768P) | T2V | 0.03 | 0.025 |
| MiniMax H3 Fast (480P) | REF2V | 0.022 | 0.018 |
| MiniMax H3 Fast (768P) | REF2V | 0.03 | 0.025 |
Partner Models
Per image
Save 10% vs vendor list| Model | Category | 512 | 1K | 2K | 4K |
|---|---|---|---|---|---|
| Nano Banana 2.1 | T2I / I2I | Vendor list price: 0.03360 USD per image.0.03024 | Vendor list price: 0.05040 USD per image.0.04536 | Vendor list price: 0.1130 USD per image.0.1017 | |
| Nano Banana 2 | T2I / I2I | Vendor list price: 0.0450 USD per image.0.0405 | Vendor list price: 0.0670 USD per image.0.0603 | Vendor list price: 0.1010 USD per image.0.0909 | Vendor list price: 0.1510 USD per image.0.1359 |
| Nano Banana Pro | T2I / I2I | Vendor list price: 0.1340 USD per image.0.1206 | Vendor list price: 0.1340 USD per image.0.1206 | Vendor list price: 0.240 USD per image.0.216 |
Per output image
(Input images on an edit: 0.003 credits each, on top of the output rate)
| Model | Category | 1K | 2K |
|---|---|---|---|
| Qwen-Image 3.0 | T2I / I2I | 0.03 | 0.03 |
| Qwen-Image 3.0 Pro | T2I / I2I | 0.04 | 0.075 |
Per second of output video
Save 10% vs vendor list| Model | Category | 720p | 1080p | 4K |
|---|---|---|---|---|
| Veo 3.1 (Standard) | T2V / I2V | Vendor list price: 0.40 USD per second.0.36 | Vendor list price: 0.40 USD per second.0.36 | Vendor list price: 0.60 USD per second.0.54 |
| Veo 3.1 (Fast) | T2V / I2V | Vendor list price: 0.10 USD per second.0.09 | Vendor list price: 0.120 USD per second.0.108 | Vendor list price: 0.30 USD per second.0.27 |
| Veo 3.1 (Lite) | T2V / I2V | Vendor list price: 0.050 USD per second.0.045 | Vendor list price: 0.080 USD per second.0.072 |
Per second of output video
Save 10% vs vendor list| Model | Category | Any resolution (720p / 1080p / 4K) |
|---|---|---|
| Avatar 4 Photo | AVATAR | Vendor list price: 0.050 USD per second.0.045 |
| Avatar 5 Digital | AVATAR | Vendor list price: 0.050 USD per second.0.045 |
Per second of output video
Save 10% vs vendor list| Model | Category | Std | Std + Audio | Pro | Pro + Audio | 4K |
|---|---|---|---|---|---|---|
| V3 | T2V / I2V | Vendor list price: 0.0840 USD per second.0.0756 | Vendor list price: 0.1260 USD per second.0.1134 | Vendor list price: 0.1120 USD per second.0.1008 | Vendor list price: 0.1680 USD per second.0.1512 | Vendor list price: 0.420 USD per second.0.378 |
| V3 | MC | Vendor list price: 0.1260 USD per second.0.1134 | Vendor list price: 0.1680 USD per second.0.1512 | |||
| V3 Omni | OMNI | Vendor list price: 0.0840 USD per second.0.0756 | Vendor list price: 0.1120 USD per second.0.1008 | Vendor list price: 0.1120 USD per second.0.1008 | Vendor list price: 0.140 USD per second.0.126 | Vendor list price: 0.420 USD per second.0.378 |
Per second of output video
| Model | Category | 480P | 720P | 720P + Audio | 1080P | 1080P + Audio |
|---|---|---|---|---|---|---|
| Wan 3.0 | T2V / I2V / REF2V | 0.05 | 0.1 | 0.1 | 0.2 | 0.2 |
| Wan 3.0 Prime | T2V / I2V / REF2V | 0.068 | 0.14 | 0.14 | 0.28 | 0.28 |
| Wan 2.7 | T2V / I2V / REF2V | 0.1 | 0.1 | 0.15 | 0.15 |
Per second of output video
| Model | Category | 480P | 720P | 1080P |
|---|---|---|---|---|
| HappyHorse 1.1 | T2V / I2V / REF2V | 0.07 | 0.14 | 0.18 |
| HappyHorse 1.0 | T2V / I2V / REF2V | 0.14 | 0.24 |
Per second of output video (24 fps, 16:9)
(ByteDance bills by video tokens: (input + output seconds) x width x height x fps / 1024, at the engine's rate per 1M tokens (Seedance 2.5 10.7). The per-second figures assume 24 fps, each engine's 16:9 frame, and no input video; a reference video adds its own seconds)
| Model | Category | 480p | 720p | 1080p |
|---|---|---|---|---|
| Seedance 2.5 | T2V / I2V | 0.103 | 0.231 |
Note: All prices reflect Nunchux's optimized inference stack. Equivalent outputs cost significantly more on standard GPU infrastructure.
Processing High Volumes?
We offer volume-based discounts.
Frequently Asked Questions
Have questions? We've got answers.
You're billed based on actual usage. Pricing is designed to be predictable and transparent.
Need more? Let's talk.
Custom SLAs as dedicated capacity, private deployment, volume discounts and a dedicated ML engineer to support your team.
Model Optimization
Compress and optimize your proprietary models using Nunchux deep tech. Reduce serving costs by up to 100x. Priced per model based on size and optimization depth.
Edge Deployment
Deploy optimized models directly on laptops and mobile devices. On-device inference with low latency and full data privacy. Pricing based on deployment scale and device targets.