Credit rates

Every model, export, compute job and platform fee, in credits. A credit is $0.10 or ₹10. Charges follow what each provider actually reports. See plans.

Language models

Chat, planning, writing the video and every check the agent runs. Charged on the tokens the provider reports, thinking included; cached input is billed at the provider’s discounted rate.

ItemInput / 1M tokensOutput / 1M tokensPer web search
Gemini 3.8 Flash Google18.7593.750.35
Claude Opus 5.5 Anthropic1005000.25
Claude Sonnet 5.5 Anthropic502500.25
Claude Haiku 5.5 Anthropic12.512.50.25
GPT-6 Astra OpenAI5001,2500.25
GPT-6.1 Sol OpenAI1002500.25
GPT-6 Luna OpenAI512.50.25

Images

Per picture at its typical quality. Reference images, thinking and edits are charged as the provider bills them; a picture the provider refuses is still charged for the work it did.

ItemPer image
Nano Banana 2 Lite Google · 1K0.84
Nano Banana 2 Google · 5121.13
Nano Banana 2 Google · 1K1.68
Nano Banana 2 Google · 2K2.53
Nano Banana 2 Google · 4K3.78
Nano Banana 2.1 Google · 1K0.84
Nano Banana 2.1 Google · 2K1.26
Nano Banana 2.1 Google · 4K2.83
Nano Banana Pro Google · 1K3.35
Nano Banana Pro Google · 2K3.35
Nano Banana Pro Google · 4K6
GPT Image 2.5 Flare OpenAI1.32
GPT Image 2.5 Sunburst OpenAI1.32

Video clips

Generated footage, per second of clip at each resolution the model offers.

ItemPer second
Veo 3.1 Lite Google · 720p1.25
Veo 3.1 Lite Google · 1080p2
Gemini Omni 1.1 Flash Google · 360p20.31
Gemini Omni 1.1 Flash Google · 720p20.31
Gemini Omni 1.1 Flash Google · 1080p20.31
Gemini Omni 1.1 Flash Google · 4k20.31

Voice, music and transcription

Narration per thousand characters of script; music and transcription per minute of audio.

ItemPer 1,000 charactersPer minute
Gemini 3.8 Flash TTS Google · voice0.39—
Gemini 3.8 Flash-Lite TTS Google · voice0.26—
Lyria 3.5 Google · music—2
Gemini 3.5 Transcribe Google · transcription—0.13
GPT Transcribe OpenAI · transcription—0.11

Video export (AWS Lambda)

Exporting a finished video to an MP4. The Export button shows the figure for your video before you press it; the charge is what AWS reports the render actually used, plus the transfer of the file. Cancelling a render that has started still pays for it, because AWS finishes it. Exporting in your own browser is free.

Item15-second video30-second video60-second video
1080p · 30 fps 0.741.432.81
1080p · 60 fps 1.382.715.36
1440p · 30 fps 0.791.533.02
1440p · 60 fps 1.452.855.65
4K · 30 fps 0.951.863.66
4K · 60 fps 1.673.296.52

Compute

The cloud machine (Google Cloud Run, 2 vCPU and 4 GB) that writes the video's code and prepares uploaded videos, charged per job by the time it ran. Every job is at least the minimum; past that, about 0.09 credits a minute. Jobs that write a video also cover the preview frames taken while writing. Poster frames are charged per frame.

ItemCredits
Preparing an uploaded video minimum per job1
Writing or revising a video or designed post minimum per job1.5
Each further minute of a job past the minimum0.09
Poster or preview frame per frame, AWS Lambda0.1

Publishing and inbox

Posting, scheduling, analytics, comments and messages are free on every platform except X, which charges for every post and message sent through its API. Those charges pass straight through. A post whose text contains a link costs X far more than one without. Boosted posts spend from your own ad account, never from credits.

ItemCredits
X post or reply 0.38
X post or reply with a link 5
X direct message 0.38
Deleting an X post or reply 0.25
Any other platform post, schedule, reply, messageFree