Model rates reference
The exact per-call rates every job is billed at — token rates for text models, per-image and per-second rates for media. Auto-generated from the pricing catalog.
Every job bills only what it actually uses, at the rates below, and returns an itemized receipt. For the plain-language version — free credits, how billing works, what typical jobs cost — see Pricing.
Text models
The agentic LLM driving a skill, billed per token.
| Model | Input / 1M tokens | Output / 1M tokens |
|---|---|---|
Claude Haiku 4.5 (anthropic/claude-haiku-4-5) | $0.25 | $1.25 |
Claude Opus 4.8 (anthropic/claude-opus-4-8) | $5.00 | $25.00 |
Claude Sonnet 4.6 (anthropic/claude-sonnet-4-6) | $3.00 | $15.00 |
Image models
Billed per generated image (or per megapixel where noted).
| Model | Rate |
|---|---|
GPT Image 2 (openai/gpt-image-2) | $0.211 / image |
| · low · 1024x768 | $0.0050/image |
| · low · 1024x1024 | $0.0060/image |
| · low · 1024x1536 | $0.0050/image |
| · low · 1920x1080 | $0.0050/image |
| · low · 2560x1440 | $0.0070/image |
| · low · 3840x2160 | $0.012/image |
| · medium · 1024x768 | $0.037/image |
| · medium · 1024x1024 | $0.053/image |
| · medium · 1024x1536 | $0.042/image |
| · medium · 1920x1080 | $0.040/image |
| · medium · 2560x1440 | $0.056/image |
| · medium · 3840x2160 | $0.101/image |
| · high · 1024x768 | $0.145/image |
| · high · 1024x1024 | $0.211/image |
| · high · 1024x1536 | $0.165/image |
| · high · 1920x1080 | $0.158/image |
| · high · 2560x1440 | $0.222/image |
| · high · 3840x2160 | $0.401/image |
Nano Banana Pro (edit) (google/nano-banana-pro-edit) | $0.150 / image |
| · 1K / 2K | $0.150/image |
| · 4K | $0.300/image |
Seedream v4 (edit) (bytedance/seedream-v4-edit) | $0.030 / image |
GPT Image 2 (edit) (openai/gpt-image-2-edit) | $0.219 / image |
| · low · 1024x768 | $0.011/image |
| · low · 1024x1024 | $0.015/image |
| · low · 1024x1536 | $0.018/image |
| · low · 1920x1080 | $0.017/image |
| · low · 2560x1440 | $0.019/image |
| · low · 3840x2160 | $0.024/image |
| · medium · 1024x768 | $0.043/image |
| · medium · 1024x1024 | $0.061/image |
| · medium · 1024x1536 | $0.054/image |
| · medium · 1920x1080 | $0.053/image |
| · medium · 2560x1440 | $0.068/image |
| · medium · 3840x2160 | $0.113/image |
| · high · 1024x768 | $0.151/image |
| · high · 1024x1024 | $0.219/image |
| · high · 1024x1536 | $0.178/image |
| · high · 1920x1080 | $0.158/image |
| · high · 2560x1440 | $0.234/image |
| · high · 3840x2160 | $0.413/image |
Imagen 4 (google/imagen-4) | $0.040 / image |
Kling v3 (image edit) (kuaishou/kling-v3-image-edit) | $0.028 / image |
Kling v3 (image) (kuaishou/kling-v3-image) | $0.028 / image |
Nano Banana Pro (google/nano-banana-pro) | $0.150 / image |
| · 1K / 2K | $0.150/image |
| · 4K | $0.300/image |
Seedream v4 (bytedance/seedream-v4) | $0.030 / image |
Video models
Billed per second of generated video.
| Model | Rate |
|---|---|
Kling v3 Pro — image→video (kuaishou/kling-v3-i2v) | $0.11 / second |
| · audio off | $0.112/s |
| · audio on | $0.168/s |
| · audio + voice control | $0.196/s |
Kling AI Avatar v2 Pro — lip-sync (kuaishou/kling-avatar-v2) | $0.12 / second |
| · lip-sync | $0.115 / second of output |
Kling O3 Pro — reference→video (kuaishou/kling-o3-r2v) | $0.11 / second |
| · audio off | $0.112/s |
| · audio on | $0.140/s |
Kling v3 Pro — text→video (kuaishou/kling-v3-t2v) | $0.11 / second |
| · audio off | $0.112/s |
| · audio on | $0.168/s |
| · audio + voice control | $0.196/s |
Seedance 2.0 Fast — image→video (bytedance/seedance-2-fast-i2v) | $0.24 / second |
Seedance 2.0 Fast — reference→video (bytedance/seedance-2-fast-r2v) | $0.24 / second |
| · without video reference | $0.242/s |
| · with video reference | $0.145/s |
Seedance 2.0 Fast — text→video (bytedance/seedance-2-fast-t2v) | $0.24 / second |
Seedance 2.0 — image→video (bytedance/seedance-2-i2v) | $0.30 / second |
| · 720p | $0.302/s |
| · 1080p | $0.682/s |
Seedance 2.0 — reference→video (bytedance/seedance-2-r2v) | $0.30 / second |
| · without video reference | $0.302/s |
| · with video reference | $0.181/s |
Seedance 2.0 — text→video (bytedance/seedance-2-t2v) | $0.30 / second |
| · 720p | $0.303/s |
| · 1080p | $0.682/s |
Veo 3 Fast — image→video (google/veo-3-fast-i2v) | $0.10 / second |
| · audio off | $0.100/s |
| · audio on | $0.150/s |
Veo 3 Fast — text→video (google/veo-3-fast-t2v) | $0.10 / second |
| · audio off | $0.100/s |
| · audio on | $0.150/s |
Veo 3 — image→video (google/veo-3-i2v) | $0.20 / second |
| · audio off | $0.200/s |
| · audio on | $0.400/s |
Veo 3 — text→video (google/veo-3-t2v) | $0.20 / second |
| · audio off | $0.200/s |
| · audio on | $0.400/s |
Veo 3.1 — reference→video (google/veo-3.1-r2v) | $0.20 / second |
| · audio off | $0.200/s |
| · audio on | $0.400/s |
Audio models
Billed per call / per 1k characters as noted.
| Model | Rate |
|---|---|
ElevenLabs v3 — expressive text→speech (elevenlabs/tts-v3) | $0.100 / 1k chars |
| · text-to-speech | $0.100 / 1,000 characters |
ElevenLabs Multilingual v2 — text→speech (elevenlabs/tts-multilingual-v2) | $0.100 / 1k chars |
| · text-to-speech | $0.100 / 1,000 characters |
ElevenLabs Scribe v2 — speech→text (elevenlabs/scribe-v2) | $0.00 / second |
| · transcription | $0.0080 / minute of audio |
| · with keyterms biasing | $0.010 / minute of audio |