Plans & Credits

Every Morphic plan includes a monthly allowance of credits. You spend them on generating images, video and audio, on replies from Copilot (also called Chat), and on AI edits in Compose. This page lists what each one costs, starting with how plans and credits work.

Subscription plansCopy Anchor Link

Compare plans and prices at morphic.com/pricing. What each plan includes is under Plans.

How credits workCopy Anchor Link

Plan credits are a monthly allowance. On monthly billing they reset on your billing date; on yearly billing they reset on the 1st of every month at 00:00 UTC. Unused plan credits do not roll over; credits from a credit pack last until the pack expires.

On plans with more than one member, credits are pooled and shared by everyone in the organization. Plan credits are used first, then any credit packs.

Studio shows what you are actually charged: the price on the generate button, 'credits consumed' on each Copilot reply, and Credit usage in Settings. The tables on this page show typical costs and can lag behind product changes. Where they differ, Studio is correct. Copilot's answers about pricing are not binding.

Results are AI-generated, so they are not 100% accurate or predictable, and getting the one you want often takes a few attempts. Each attempt is a new run, so each one uses credits. Credits used on a finished run are not returned; a failed generation, or one stopped before the model provider starts it, gets its credits back automatically. See Pricing FAQs.

Free planCopy Anchor Link

Everyone starts on the Free plan when they sign up. No credit card is required. Use it to try Morphic before you choose a plan. It includes:

Adaptive pricingCopy Anchor Link

Rates marked * depend on your plan, so the tables show a range wherever your plan changes the number. The ranges cover the paid plans on both billing periods, monthly and annual, including the earlier Basic price that existing subscribers keep; on free, education and contract accounts the rate can differ. The generate button shows your exact price before you run, except on speech priced per character; see Text to speech.

Copilot (Chat)Copy Anchor Link

Copilot is also called Chat in Studio: the 'Copilot' tab and the 'Chat' button open the same assistant, and its replies are charged the same way in both. See Copilot and chats.

Copilot's replies use credits, on top of anything they generate. That includes one-word messages like "go" or "retry": before every reply, Copilot reads the chat, so the same message costs more in a long chat. A model you pick is always charged; Auto can be too, depending on your account.

What changes the costCopy Anchor Link

• Chat length. The longer the conversation, the more each reply reads.
• Steps. A reply that plans, starts a generation and checks the result takes several steps, and each one reads the chat again.
• Switching models. The new model reads the chat from the start, so its first reply costs more. Morphic warns you before you switch.
• Breaks. When a chat has been idle for a while, up to about an hour depending on the model, the next reply reads it from the start, as after a switch.

Copilot reply costsCopy Anchor Link

Typical credits for one reply, by the model the chat runs on:

Reduce Copilot costsCopy Anchor Link

• Start a new chat for new work. It starts without the old history, so each reply reads far less. 'Continue in new chat' carries the history over, so its replies cost the same as the original's.
• Stay on Auto for everyday back-and-forth.
• Pick a model at the start of a chat rather than partway through a long one.
• Put the whole request in one message. Four short follow-ups are four replies.

Credits used in a chatCopy Anchor Link

1. In the 'Chats' sidebar of a file, right-click the chat
2. Click 'Total credits'

'Credits used in this chat' shows what the chat was charged, what was refunded, and the net, split by type. Copilot's replies are under 'LLM'. To compare chats, open Credit usage and set 'Group by' to Chats.

Tools inside a chatCopy Anchor Link

Some of what Copilot does carries its own cost, whichever model the chat is on. Each one appears in the chat's credit breakdown.

ComposeCopy Anchor Link

What is chargedCopy Anchor Link

When you ask Copilot to work on the timeline, the 'credits consumed' figure above its reply is the sum of these charges. Most requests use only the first two.

Server exportCopy Anchor Link

In the Export dialog, 'Render on' picks where the video renders. 'This device' renders in your browser and is free. 'Server', on paid plans, is charged for the time the render takes on Morphic's servers, plus about 30 seconds to start up, with a one-minute minimum. Copilot always exports this way.

Render time is how long the server works, not the length of the video. Starting a server export needs at least 45 credits, which are set aside and then settled against the actual cost. If a server export fails on Morphic's side, the credits come back.

Transcript costsCopy Anchor Link

Removing pauses or filler words, adding subtitles, and ducking music under speech all start from a transcript, so Copilot knows where each word falls. It is charged in two parts:

• The transcription, per second of speech at the transcription model's rate, with a minimum of 1 credit. At ElevenLabs Scribe v2 rates, anything under about a minute and a half costs the 1-credit minimum, and a 10-minute interview about 6 to 8. The rates for every model are under Transcription, and depend on your plan; see Adaptive pricing.
• The background work that runs it, usually 1 to 2 credits.

Copilot makes the transcript once and reuses it. Removing pauses, then filler words, then adding subtitles on the same clip pays for one transcript, not three.

The 'Captions' tab is separate from Copilot: 'Generate captions' is charged at the transcription rate for the audio it listens to, with no steps or reply.

Edit costsCopy Anchor Link

Each step below is 1 credit.

A worked example. 'Put these three clips on the timeline with a cross dissolve between each, then add subtitles' on an empty timeline comes to:

For one change you can see and click, doing it yourself is free and quicker. See Edit with Copilot for what you can ask for.

ImageCopy Anchor Link

Text to ImageCopy Anchor Link

* Depends on your plan; see Adaptive pricing.

Ideogram 4.0 is priced by rendering speed: ~6 on Turbo (the default), ~11 on Balanced and ~18 on Quality.

Recraft V4.1 Flash renders about one megapixel at the aspect ratio you pick, so it has no resolution setting.

Flux 3 in these image tables is the image model. The Flux 3 video model shares its name and is priced under Video.

GPT Image 2.5 Flare and Sunburst share one rate card, and both charge by quality as well as resolution. The table above shows Medium, the default. Per image, by tier:

* Depends on your plan; see Adaptive pricing.

GPT Image 2 charges by quality as well as resolution. The table above shows Medium, the default. Per image, by tier:

* Depends on your plan; see Adaptive pricing.

Qwen Image 2.1 renders at the dimensions you ask for and bills per megapixel of output, so the price follows the size: about 3 to 4 credits for a 1024x1024 image, 6 to 8 for a 1080p 16:9 frame, and 24 to 29 at 4K.

Image to ImageCopy Anchor Link

* Depends on your plan; see Adaptive pricing.

Ideogram 4.0 is priced by rendering speed: ~6 on Turbo (the default), ~11 on Balanced and ~18 on Quality.

Flux 3 takes up to 10 reference images, included in the price per image.

GPT Image 2.5 Flare and Sunburst share one rate card, and both charge by quality as well as resolution. The table above shows Medium, the default. Per image, by tier:

* Depends on your plan; see Adaptive pricing.

On GPT Image 2.5, every reference image you attach adds about 2 to 3 credits; none are included, and you can attach up to 10.

GPT Image 2 charges by quality as well as resolution. The table above shows Medium, the default. Per image, by tier:

* Depends on your plan; see Adaptive pricing.

Qwen Image 2.1 bills per megapixel of output: about 6 to 7 credits for a 1024x1024 image, 11 to 13 for a 1080p 16:9 frame, and 43 to 52 at 4K. Each reference image adds about 3 to 4 credits; none are included, and you can attach up to 10.

Seedream 5.0 Pro layer separation bills per output image rather than per generation: ~4 each at 1080p and ~8 each at 2K, counting the base image it returns alongside the separated layers.

VideoCopy Anchor Link

Seedance 2.0 Fast promotion. Until October 7, 2026, Seedance 2.0 Fast is discounted on every operation. The tables below show the standard rate; while the promotion runs you pay 7 to 8 credits per second instead of 8 to 10 at 480p, and 14 to 17 instead of 17 to 21 at 720p. An attached video is billed at 4 to 5 per second instead of 5 to 6 at 480p, and 8 to 10 instead of 10 to 13 at 720p. The generate button shows the discounted price.

Clip lengths

Text to VideoCopy Anchor Link

* Depends on your plan; see Adaptive pricing.

Seedance 2.5 Draft is Seedance 2.5 at 480p only, so a 5-second draft is 72 to 87 credits. A draft stays available for 7 days, and rendering its 1080p final is charged separately at the Seedance 2.5 1080p rate on the draft's own length: 398 to 481 credits for that same 5-second clip.

Image to VideoCopy Anchor Link

* Depends on your plan; see Adaptive pricing.

On Seedance 2.5 and the MiniMax H3 models, your input image sets the aspect ratio and is not charged.

Legacy pricing for Seedance 1.5Copy Anchor Link

Frames to VideoCopy Anchor Link

* Depends on your plan; see Adaptive pricing.

Your frames are never charged. Flux 3 takes 2 to 10 frames: two set a first-to-last transition, more become keyframes spaced along the clip.

Reference to VideoCopy Anchor Link

* Depends on your plan; see Adaptive pricing.

On this operation MiniMax H3 Max is billed at one rate for both 480p and 768p. LTX 2.5 is audio-driven on this operation: it animates a still from a reference audio clip at a flat 15/s.

Reference costsCopy Anchor Link

Most models take the files you attach at no extra cost, so you pay only for the clip you generate. These charge for references:

* Depends on your plan; see Adaptive pricing.

Kling O3, Gemini Omni Flash, Gemini Omni Flash 1.1, Wan 3.0 and Wan 3.0 Prime take reference video at no extra cost.

Seedance video input rateCopy Anchor Link

Give Seedance a video, as a reference, a clip to edit or a lip-sync source, and the whole generation switches to the video input rate: about 40% below the normal rate, billed on the length of every input video plus the clip you generate. A video under 4 seconds counts as 4.

* Depends on your plan; see Adaptive pricing.

Three 5-second generations at 720p on Seedance 2.5:

On a 10-second clip at 720p the same comparison is 330-400 credits from images or audio only, 280-336 with a 4-second video and 400-480 with a 10-second one.

Seedance 2.5 and 2.5 Draft take up to 50 references (30 images, 10 videos, 10 audio clips); Seedance 2.0 takes 9 images, 3 videos and 3 audio clips.

MiniMax H3Copy Anchor Link

A reference video is billed on its actual length at the same rate as the clip: a 5-second clip at 768p with a 3-second reference video is billed on 8 seconds, so 112-136* credits. You can attach 9 images, 3 videos and 3 audio clips.

MiniMax H3 MaxCopy Anchor Link

References share a free allowance of about 4 megapixels of images; reference audio always fits in it. A reference video is billed by its length and the resolution you generate at, on top of the clip's 14-17*/s (28-34*/s at 1080p):

* Depends on your plan; see Adaptive pricing.

These are for a video on its own; images take from the same allowance. Image size counts, so one 12-megapixel photo adds about 26-32* credits. You can attach 9 images, 3 videos and 3 audio clips, 12 files in total.

Video to VideoCopy Anchor Link

* Depends on your plan; see Adaptive pricing.

A Seedance edit is billed on the source clip as well as the output. Every edit carries a source video, so these are already the video input rates. Billed seconds = the source clip + what Seedance generates, with the source counting as at least 4 seconds. A 5-second edit at 720p on Seedance 2.5 is billed on 10 seconds, so 200 to 240 credits. Further reference videos add their length on top. On an edit the output matches the input length and the source must run 4 to 30 seconds; on an extend you set how long the new footage runs.

Gemini Omni Flash 1.1 also outputs 360p at 5/s on an edit.

Flux 3 extends an existing clip (source up to 15 seconds and 50 MB). You are billed on the video it generates, not on the input.

Veed Subtitles is billed on the input video's resolution, with a 1-minute (32-credit) minimum per generation.

Lip syncCopy Anchor Link

* Depends on your plan; see Adaptive pricing.

The Seedance rates above are for syncing a still image. Driving the sync from a video instead switches to the video input rate, billed on the source clip plus the output.

AudioCopy Anchor Link

Audio is billed per second of generated output (/s) unless the table shows another unit. Flat prices are charged once per generation, speech to speech bills the recording you upload, and text to speech and dialogue on ElevenLabs, Inworld and xAI bill per character of your text (/1K chars). Per-second models charge the listed minimum upfront and settle the balance once the final length is known, so the exact cost is confirmed after you generate. Each operation below lists the models available and what they’re best suited for.

Text to speechCopy Anchor Link

* Depends on your plan; see Adaptive pricing.

/1K chars = credits per 1,000 characters of your text. Models priced this way charge the whole cost when you start: characters ÷ 1,000 × the rate, rounded up once to a whole credit, with a 5-credit minimum. Nothing is settled when the audio finishes.

• Every character counts, including spaces, punctuation and audio tags such as [laughs].
• About 1,000 characters make a minute of speech. That costs 14 to 17 credits on ElevenLabs v4, 7 to 9 on ElevenLabs v4 Turbo, 5 to 6 on Inworld TTS 2, and the 5-credit minimum on Inworld TTS 2 Flash and xAI Grok TTS.
• The generate button shows the 5-credit minimum for per-character models, not the full charge. Credit usage in Settings shows what each run cost.

Gemini 3.8 promotion. Until January 1, 2027, Gemini 3.8 Flash TTS and Gemini 3.8 Flash-Lite TTS are half price in text to speech and text to dialogue. The tables show the standard rate; while the promotion runs you pay about 0.05 to 0.06 credits per second on Gemini 3.8 Flash TTS and 0.03 to 0.04 on Gemini 3.8 Flash-Lite TTS, so a minute of speech comes to the 5-credit minimum on both, and a two-minute read is 7 to 8 credits on Flash and 5 on Flash-Lite.

Text to dialogueCopy Anchor Link

* Depends on your plan; see Adaptive pricing.

ElevenLabs v4 and ElevenLabs v4 Turbo dialogue is charged per character, like text to speech, on the text of all its lines counted together. Speaker names are not counted. Gemini 3.8 Flash TTS, Gemini 3.8 Flash-Lite TTS and Seed Audio 1.0 dialogue are charged per second of the finished audio, like their text to speech, and the Gemini 3.8 promotion above applies to dialogue too.

Text to musicCopy Anchor Link

* Depends on your plan; see Adaptive pricing.

Text to sound effectsCopy Anchor Link

* Depends on your plan; see Adaptive pricing.

Speech to speechCopy Anchor Link

What you pay depends on how long the recording you upload is, not on the audio you get back. You are charged when you start the conversion, and the exact number is shown on the button first, so there is no estimate to wait on:

Every conversion is charged on its own, so trying the same recording on a second voice costs the same again. A conversion never costs less than 5 credits.

TranscriptionCopy Anchor Link

* Depends on your plan; see Adaptive pricing.

Transcription is charged on the media you feed in rather than the text it returns. If you ask for a section of a file, you pay for that section rather than the whole thing. A transcription never costs less than 1 credit, which covers roughly five to six minutes of Whisper. When Copilot makes a transcript for an edit in Compose, the background work that runs it adds 1 to 2 credits; see How the transcript is charged.

Note: /s = credits per second of generated audio. Flat prices are charged once per generation regardless of length. For per-second models the minimum is charged upfront and the balance settled from the actual output length: a 10-second Seed Audio 1.0 clip is about 6 to 7 credits and a minute of Gemini 3.8 Flash TTS about 7 to 8 at the standard rate, while a Seed Audio clip under about 7 seconds, or a Gemini 3.8 Flash TTS clip under about 40 seconds, is billed at the 5-credit minimum.

Stem separationCopy Anchor Link

* Depends on your plan; see Adaptive pricing.

Stem separation is charged on the track you feed in rather than on the stems it returns, so the six-stem split costs more per minute but not more per stem. A three-minute track comes to about 39 to 48 credits split two ways and about 78 to 96 split six ways. One track goes in, up to 10 minutes long, and every stem comes back as its own audio file.

Upscale and enhanceCopy Anchor Link

Upscale videoCopy Anchor Link

Credits = rate × seconds of video. The rate depends on the size your video comes out at, which is your clip's size times the upscale factor. A 720p clip at 2x comes out at 1440p, and a 1080p clip at 2x comes out at 4K.

* Depends on your plan; see Adaptive pricing.

Example: a 24-second 1080p clip upscaled 2x on Topaz Astra 2 comes out at 4K, so it costs 24 × 70-85 = 1,680 to 2,040 credits.

• The frame rate you pick does not change the price, except on Crystal Video: footage above 30 fps costs twice the rate, and above 60 fps three times. Crystal Video's rates are for a 16:9 or 9:16 frame.
• Seedance 2.5 Final renders the 1080p final of a Seedance 2.5 Draft made in the last 7 days.

Upscale imageCopy Anchor Link

Credits = rate × megapixels of the upscaled image, rounded up. Megapixels = width × height ÷ 1,000,000.

* Depends on your plan; see Adaptive pricing.

Example: a 1024 × 1024 image upscaled 2x becomes 2048 × 2048, about 4.2 megapixels. On Topaz Wonder 3.5 that costs 6 to 8 credits.

Enhance and editCopy Anchor Link

Credit packsCopy Anchor Link

If you run out of plan credits, top up at https://studio.morphic.com/settings/billing/credits

	Growth	Pro	Plus	Ultra
Price	$25	$100	$250	$500
Credits	2500	11110	29400	62500
Validity	180 days	180 days	180 days	180 days
Subscription Required	Required	Required	Required	Required

How packs workCopy Anchor Link

• Available only to paid users (some packs may be restricted to certain plans).
• Expiration starts from the purchase date, regardless of when credits are first used.
• Credit packs are non-refundable and cannot be paused once purchased.
• Credits roll over until expiration, unlike subscription credits, which reset each month.
• You can activate multiple credit packs at once; credits will be used in order of expiry.
• Only organization admins can purchase credit packs.
• You need an active subscription to use credit packs.

More billing questions are answered in Pricing FAQs.

---

If you have any trouble or have any questions, feel free to reach out. We're happy to help. Write to us at support@morphic.com.