MiniMax Max vs MiniMax Ultra Plan Comparison
This page compares MiniMax Max vs MiniMax Ultra side by side across pricing, usage quotas, concurrency, supported models, availability, and intended coding workflows. Use the structured comparison to choose a plan that fits your budget and usage pattern, and confirm the latest provider terms before subscribing.
| Metric | MiniMaxMaxIn stockStrong models | MiniMaxUltraIn stockStrong models |
|---|---|---|
| Provider / Type | MiniMax MiniMax Token Plan Token Plan | MiniMax MiniMax Token Plan Token Plan |
| Audience | Personal / Enterprise | Personal / Enterprise |
| Availability | Limited | New sales stopped in October 2026. Subscriptions with auto-renew stay on the old price and benefits. Others end when the current period ends. | Limited | New sales stopped in October 2026. Subscriptions with auto-renew stay on the old price and benefits. Others end when the current period ends. |
| Monthly 路 CNY | 楼119 / Monthly | 楼469 / Monthly |
| Request quota (5h / week / month) | 4,500 (Estimated) / 45,000 (Estimated) / 180,000 (Estimated) | 15,000 (Estimated) / 150,000 (Estimated) / 600,000 (Estimated) |
| Token quota (5h / week / month) | - / - / 1,800,000,000 (Estimated) | - / - / 7,100,000,000 (Estimated) |
| Agent concurrency | 4-5 | 6-7 |
| Supported models | ||
| Harness / client tools | - | - |
| Overview and notes | Closed to new purchases. Renewing subscriptions use M3.1 Flash for text and do not include H3. The M2.5 and M2.7 names route to M3.1 Flash. The M3 monthly programming estimate assumes approximately 50K tokens per request. M2.7, M2.7-highspeed, and M3 share a quota. The 5-hour figures estimate exclusive use of the corresponding model. Plus does not include video generation. Max and Ultra include 3 and 5 videos per day respectively. Shared multimodal quota: Token Plan covers flagship M3 / M2.7 / image / speech models under one shared usage meter; a few special models such as H3, voice design, and quick voice clone are not supported. New sales stopped in October 2026. Subscriptions with auto-renew stay on the old price and benefits. Others end when the current period ends. | Closed to new purchases. Renewing subscriptions use M3.1 Flash for text and do not include H3. The M2.5 and M2.7 names route to M3.1 Flash. The M3 monthly programming estimate assumes approximately 50K tokens per request. M2.7, M2.7-highspeed, and M3 share a quota. The 5-hour figures estimate exclusive use of the corresponding model. Plus does not include video generation. Max and Ultra include 3 and 5 videos per day respectively. Shared multimodal quota: Token Plan covers flagship M3 / M2.7 / image / speech models under one shared usage meter; a few special models such as H3, voice design, and quick voice clone are not supported. New sales stopped in October 2026. Subscriptions with auto-renew stay on the old price and benefits. Others end when the current period ends. |
| Official link | Open provider | Open provider |
Pricing
MiniMaxMax
楼119 / Monthly
MiniMaxUltra
楼469 / Monthly
Core info
MiniMaxMax
AvailabilityLimited | New sales stopped in October 2026. Subscriptions with auto-renew stay on the old price and benefits. Others end when the current period ends.
Agent concurrency4-5
MiniMaxUltra
AvailabilityLimited | New sales stopped in October 2026. Subscriptions with auto-renew stay on the old price and benefits. Others end when the current period ends.
Agent concurrency6-7
Quota
MiniMaxMax
Request quota4,500 (Estimated) / 45,000 (Estimated) / 180,000 (Estimated)
Token quota- / - / 1,800,000,000 (Estimated)
MiniMaxUltra
Request quota15,000 (Estimated) / 150,000 (Estimated) / 600,000 (Estimated)
Token quota- / - / 7,100,000,000 (Estimated)
Supported models
MiniMaxMax
MiniMaxUltra