Leonardo.Ai vs Elyumfor MiniMax H3
Elyum previews MiniMax H3 for free and charges only a keep, while Leonardo.Ai deducts tokens on generate and does not publish the H3 rate.
MiniMax H3 is MiniMax's video model for 5 to 15 second clips from text, a still, or reference images and audio. The H3 look holds from 480p through 4K on Elyum. You run one SKU as text to video, image to video, or reference to video. Rate is per second, so a 5s clip and a 15s clip are the same model at different lengths. Practitioners pick it when they want that look without changing vendors mid-shot.
MiniMax H3 is the same vendor model on both platforms. The bill, the bad-take rule, and the extras change. Leonardo.Ai sells Fast Tokens from Essential at $12/mo. Third-party video always consumes tokens, even on Ultimate at $60/mo, and the H3 rate is not published. Elyum lists H3 on Elyum pricing in credits (1 credit = $0.01), previews every render free, and charges only a keep. Leonardo.Ai also ships a canvas, first-party models, team seats, and a rollover token bank. Elyum has none of those, and no timeline editor, stock footage, avatars, mobile app, or community feed.
Leonardo.Ai is the better desk if H3 sits next to Lucid Origin, Phoenix, and canvas edits in one subscription. Elyum is the better meter if H3 is the job and you want to see the take before credits move. A longer write-up of the full stack is on the Leonardo.Ai alternative page. This comparison will not invent an H3 token number Leonardo.Ai did not publish.
MiniMax H3 on Leonardo.Ai
MiniMax H3 is listed among Leonardo.Ai's newer third-party video models added in 2026. Leonardo.Ai does not publish the MiniMax H3 token cost. A fetch of Leonardo.Ai pricing returned HTTP 403; third-party guides also left H3 unverified. Web-app Fast Tokens: Essential $12/mo (8,500 plus a 25,500 rollover bank), Premium $30/mo, Ultimate $60/mo. The research does not name a plan lock for H3. Relaxed generation on Premium and Ultimate is token-free only on first-party models. Third-party video still burns tokens at full rate on Ultimate.
Tokens deduct when you generate. Support says failed jobs should restore tokens, but restoration is not necessarily instantaneous. Adobe community and Trustpilot reports from 2025-2026 describe tokens still gone after a failed or off-instruction render. Manual refund: email [email protected] with generation ID and timestamp. Paid plans get private generations and full IP ownership. Free tier is 150 tokens/day with no rollover. Outputs are public by default and carry a watermark. Use is non-exclusive. H3 max resolution and clip length on Leonardo.Ai were not independently verified.
MiniMax H3 on Elyum
On Elyum, MiniMax H3 runs 5 to 15 seconds as text to video, image to video, or reference to video, from 480p through 4K. Rate is per second: 5 credits/s at 480p, 7 at 768p, 15 at 2K, 18 at 4K. A 5s keep is 25 credits ($0.25) at 480p, 35 credits ($0.35) at 768p, 75 credits ($0.75) at 2K, and 90 credits ($0.90) at 4K. A 15s 4K keep is 270 credits ($2.70).
Every render comes back as a free preview. Credits move only when you keep. Kill the rest and the refund is automatic and in full. Failed generations are never charged. Free is up to 150 credits once (our algorithm may grant them) (one kill a month). Plus is $9/mo, Pro $29/mo, Ultra $69/mo, and paid plans kill freely. Kept files download clean, no watermark, commercial use included. Same studio from Claude, ChatGPT, or Cursor via the MCP server or the REST API.
What a MiniMax H3 take costs
| Scenario | Leonardo.Ai | Elyum |
|---|---|---|
| One 5s take you keep | Not published; Fast Tokens deduct on generate | 25 credits ($0.25) at 480p; 90 credits ($0.90) at 4K |
| Ten takes, keep two | Not published; tokens on each generate, no free preview | 2 × 25 credits = 50 credits ($0.50) |
| A failed generation | Tokens deducted; restore not always instant; email support with ID | Never charged |
| Monthly plan needed to access the model | Not published as H3-gated; Fast Tokens from $12/mo Essential | Free plan, up to 150 credits once (our algorithm may grant them), no card |
| Watermark on the file | Yes on free web-app generations; no on paid | None on kept takes |
| Unlimited-style generation | Relaxed on Premium/Ultimate, first-party only; H3 still tokens | No unlimited H3; listed per-second keep price |
Elyum: 1 credit = $0.01, charged only when you keep a take. Leonardo.Ai figures from its own pricing pages and independent reviews, verified 2026-09-07; sources at the end. Full rate card on the MiniMax H3 page.
Where the two differ
Leonardo.Ai uses Fast Tokens on the web app and a separate dollar API. MiniMax H3 token cost is not published. Elyum uses credits (1 credit = $0.01) with listed per-second H3 rates. Listed Elyum prices already sit about 20% under Higgsfield, OpenArt, and Runway anchors.
Leonardo.Ai deducts on generate. Failed jobs should auto-restore, but restoration is not necessarily instantaneous. Users report emailing [email protected] with generation ID and timestamp. On Elyum, preview is free, a keep is the only charge, and a kill refunds in full. Failed generations are never charged.
Elyum runs MiniMax H3 from 480p through 4K, clips 5 to 15 seconds. Leonardo.Ai's H3 max resolution and clip length were not independently verified. Do not assume Leonardo.Ai matches Elyum's 4K cap.
Leonardo.Ai is stronger if you want canvas editing, first-party models (Lucid Origin, Phoenix, FLUX, Motion), team seats from Team Starter at $72/mo, and public free-tier sharing. Elyum has no timeline editor, stock footage, avatars, mobile app, team seats, or community feed.
Leonardo.Ai API is separate from web Fast Token plans. Developers buy dedicated API credits and a bearer-token key. Usage is billed in dollars. Elyum MCP and REST API use the same studio credit balance as the web app.
Leonardo.Ai free-tier generations are public by default and non-exclusive. Leonardo may reuse or display them. Paid subscribers get private generations and full IP ownership. Elyum kept takes include commercial use and no watermark.
Pick Leonardo.Ai when
- You already use Leonardo.Ai's canvas and first-party models (Lucid Origin, Phoenix, Motion) on the same Fast Token subscription.
- You need team seats and a shared token pool (Team Starter from $72/mo, three-seat minimum).
- You want unused monthly tokens parked in the rollover bank rather than lost at the month boundary.
- You want one Leonardo.Ai login for Veo 3.1, Kling 3.0 Turbo, Wan 3.0, Seedance, and MiniMax H3.
Pick Elyum when
- Preview is free. Credits move only on a keep. A kill refunds in full.
- H3 rates are listed per second through 4K. A 5s 480p keep is 25 credits ($0.25).
- Free plan is up to 150 credits once (our algorithm may grant them), no card. Paid kills are unlimited.
- Call the same balance from Claude, ChatGPT, or Cursor. Other H3 desks: Higgsfield vs Elyum.
MiniMax H3 in three styles, rendered on Elyum
Getting the most out of MiniMax H3
- Write the clip as a 5 to 15 second action with a start, a move, and an end. H3 is a short clip model, not a scene pack. If the beat needs 20 seconds, cut it into two prompts.
- For image to video, treat the still as frame one. Name the camera move once (slow push-in, handheld drift) and the motion you want in the subject. Two camera verbs in one prompt fight each other.
- Reference to video is for identity, wardrobe, or a board you want held. Use a small set of consistent refs. Conflicting faces or lighting in the ref stack will smear.
- If you pass audio as a reference, describe the picture that should lock to it. Do not ask H3 to score a timeline edit. Elyum has no timeline editor, and H3 will not assemble cuts.
- Pick resolution for the keep, not the scout. Preview on Elyum is free at the size you request. A 5s 4K keep is 90 credits ($0.90). Scout at 480p if you are still hunting the take.
- Name lighting and lens in concrete terms (soft window key, 35mm, shallow focus). Avoid abstract mood words with no camera or physics. H3 responds to visible motion more than to adjectives.
Frequently asked questions
How much does a MiniMax H3 video cost on Elyum vs Leonardo.Ai?
Is MiniMax H3 the same model on both?
What happens to a bad MiniMax H3 take?
Do I need a subscription to use MiniMax H3 on Elyum?
Does Leonardo.Ai watermark MiniMax H3 output?
Can MiniMax H3 make 30-second clips?
- Leonardo.Ai Pricing (search-engine summary; direct fetch returned HTTP 403) (accessed 2026-09-07)
- Leonardo.AI pricing 2026: plans, tokens, and the real cost — eesel AI (accessed 2026-09-07)
- AI Video Generation Models: What to Use & When — Leonardo.Ai (accessed 2026-09-07)
- Veo 3 AI Video Generator – Try Now — Leonardo.Ai (accessed 2026-09-07)
- video not generated and credits gone — Adobe Community bug report (accessed 2026-09-07)
- Leonardo.Ai Reviews — Trustpilot (accessed 2026-09-07)
- Leonardo.Ai API getting-started docs (authentication and separate API billing) (accessed 2026-09-07)
- Can You Use AI-Generated Images Commercially? — Leonardo.Ai (accessed 2026-09-07)
Related comparisons
Create everything. Keep the best.
MiniMax H3 is a AI video generator you can preview free on Elyum. Keep the takes you love, kill the rest for a full refund.