Cartesia overage $38-$65 per million credits vs Sume TTS price
Cartesia charges overage of $65, $45 or $38 per million credits by plan. Sume TTS lists $47.50 per million characters. Where each one wins, with the math.
Two ways to buy the same unit
Cartesia's pricing page says one credit equals one character, so its credit prices are character prices. It lists overage per 1M credits at $65 on Pro, $45 on Startup and $38 on Scale. Sume TTS lists $0.0475 per 1,000 characters, which is $47.50 per million.
The difference is structure. Cartesia sells a monthly plan with an included credit pool, then overage. Sume bills per character with no plan to size.
Side by side
| Option | Monthly fee | Included credits | Overage per 1M |
|---|---|---|---|
| Cartesia Pro | $5 | 100K | $65 |
| Cartesia Startup | $49 | 1.25M | $45 |
| Cartesia Scale | $299 | 8M | $38 |
| Sume TTS (list) | none | none | $47.50 flat |
Where each makes sense
For light use, a flat rate wins because you pay nothing for an unused pool. If you generate 100,000 characters in a month, Sume costs $4.75. Cartesia Pro's $5 fee includes that 100K credits, so they land close together.
At volume the plan fees matter. Startup includes 1.25M credits for $49, which is $39.20 per million effective if you use the whole pool. That is below $47.50. Scale at $299 for 8M is $37.38 per million. So a team that reliably uses a full pool on a vendor plan pays less per character than Sume's list rate.
Sume's side of the trade is what comes with the call: job envelope, Idempotency-Key, polling, webhooks, word timestamps, and the same account used for captions, music and video.
Questions to ask before you switch
- Will you use the whole included pool every month? Unused credits are wasted spend on any plan.
- Do you need Sume's job and webhook handling, or do you already run your own?
- Do you need the 8000-48000 Hz sample-rate and mp3/wav/raw container choices on Sume TTS 1.0, or do you call a Sonic model directly?
- Which model do you need? The Sume TTS Router lists sonic-3.6, sonic-3.5, sonic-3, sonic-latest and sonic-preview.
A break-even sketch
Take Startup: $49 buys 1.25M credits, then overage is $45 per million. At 2M characters in a month, Cartesia Startup costs $49 plus 0.75M x $45 = $82.75, while a flat $47.50 per million is $95. At 500,000 characters, Startup is still $49 and the flat rate is $23.75. So the crossover sits near 1.03M characters for Startup versus a flat rate.
Those numbers use only the prices on the two pages and ignore taxes, credits you may carry over, and any plan change. Redo the sum with your volumes.
Takeaway
Per character, a fully used Cartesia Startup or Scale pool undercuts $47.50 per million, and Sume's flat price beats overage and light use. Compute your own monthly characters first, then compare the two lines. Plan prices were read on 2026-10-03 and can change, so recheck the vendor page before you commit.
Sources
Related posts
More in Comparisons
- Claude Code mod vs MCP server vs skill vs hook: where Sume fits
Claude Code mods, MCP servers, skills and settings hooks overlap. Which one gives an agent Sume's image and video tools, and which one only guards them.
- Colossyan 20-30 minute video cap vs Sume avatar videos of 4-60 seconds
Colossyan's pricing page lists a per-video limit of 20 to 50 minutes and 40 scenes. Sume's avatar video takes 4 to 60 seconds per job.
- Comfy Agent Ask or Auto mode vs unattended Sume Format runs
Comfy Agent asks before each run or runs on its own. Sume API runs never ask. See which controls replace the approval prompt when a batch goes overnight.
- Comfy Agent vs an MCP agent for image and video work
Comfy Agent builds and runs ComfyUI graphs for you in Comfy Cloud; an MCP agent calls hosted tools like Sume's. What differs in control, billing and location.
Written by Sume