About this vendor
Vendor-level selection advice by Y-API. The per-model pages linked below carry the model-specific checks; nothing here is a measured comparison.
Where this vendor fits
As one of the catalog’s free entries, M2.7 is a zero-cost candidate for text workloads that fit comfortably in 200k tokens: incident triage, plan drafting, tool-driven maintenance scripts. M3 is the paid upgrade when inputs outgrow that window or need media.
What to watch for
Free pricing says nothing about capacity or permanence, and the shorter window forces architecture decisions earlier: retrieval or summarization becomes mandatory where the million-token entries just fit the document in. M3 is a different model, not a paid switch on M2.7.
Models on Y-API
Prices are account-credit snapshots from the Y-API catalog. Each model links to its own guide with a cost estimator and a request example.
| Model | Context window | Input / 1M tokens | Output / 1M tokens |
|---|---|---|---|
MiniMax M2.7minimax/minimax-m2.7 | 204,800 | Free — no credit deducted | |
MiniMax M3minimax/minimax-m3 | 1,048,576 | $0.30$0.03 | $1.20$0.12 |
Account credit / Cash equivalent · $1 paid adds $10 of credit; the cash equivalent is credit consumed ÷ 10. These are Y-API prices, not the publisher’s list prices.
Choosing within the lineup
Run text work on free M2.7 while it fits. Promote to M3 when inputs routinely exceed roughly 150k tokens or when video and image input is the point — and compare M3 against Qwen3.8 Flash on the same multimodal inputs before settling.
Before you choose
Is MiniMax M2.7 really free on Y-API?
The current catalog marks it free: calls do not deduct credit. That is today’s snapshot, not a promise of unlimited capacity or permanent availability, and the upstream can withdraw free status at any time.
Sources & scope
Vendor scope on this site is exactly the reviewed model guides listed above; a vendor page exists only while at least one of its models is reviewed. Claims about architecture or positioning follow the cited pages. We do not publish measured latency, uptime or benchmark scores for any vendor.