NewModelHubCalculate Your Cost
Enter uncached and cached input separately. If your usage report gives total input plus a cached subset, subtract cached from total before filling uncached input.
Your Estimate
Cost per request: $0.000390
Estimated monthly cost: $78.00
MiniMax-M3.1-Flash Rate Card (per 1M tokens)
| Token type | Per 1M tokens |
|---|---|
| Input (uncached) | $0.15 |
| Output | $0.60 |
| Prompt caching Read (cached input) | $0.015 |
Rates from MiniMax’s official Pricing Overview (platform.minimax.io/docs/pricing/overview), verified October 24, 2026 (Asia/Shanghai). Model ID: MiniMax-M3.1-Flash. Flat pricing — no long-context surcharge on the Flash row. Flash does not use the M3 Priority / service_tier 1.5× table. Context window: 1,000,000 tokens. OpenAI-compatible base URL https://api.minimax.io/v1 ; Anthropic-compatible https://api.minimax.io/anthropic .
What Drives Your Bill
- Output is 4× input. At $0.60 vs $0.15 per million, output length is the main lever — cap max_tokens and keep answers short by instruction.
- Cache reads are 10× cheaper. Prompt-cache read is $0.015 vs $0.15 fresh input. Put stable system prompts, tool schemas, and reused context where MiniMax can hit cache (prefix match; caching applies from ≥512 input tokens per MiniMax docs).
- Do not double-count input. Uncached + cached must partition the billed input — pasting total input into both fields inflates the estimate.
- Flat rate across the 1M window. Unlike MiniMax-M3’s ≤/>512k bands, Flash’s published row is one rate for the whole context.
- API ≠ M Plan ≠ MiniMax Code. This calculator only models metered developer API tokens. Subscription / Code products have separate terms — do not assume they include API balance.
Related tools
FAQ
How much does MiniMax-M3.1-Flash cost?
Official MiniMax API list price is $0.15 per million input tokens, $0.60 per million output tokens, and $0.015 per million prompt-cache read (cached input) tokens.
What is the model ID string?
Use MiniMax-M3.1-Flash exactly. Informal spellings like “MiniMax M3 1 Flash” or “MiniMax 3.1 Flash” refer to the same model.
How should I enter cached tokens?
Enter only the cache-hit / prompt-cache-read portion in the cached field, and only the remainder in uncached input. On OpenAI-compatible responses, cached tokens often appear under usage.prompt_tokens_details.cached_tokens; on Anthropic-compatible, under usage.cache_read_input_tokens.
Does this page cover self-hosted or open weights?
No. MiniMax states that model weights will be released later. This calculator is for the cloud API rates only.
Is this an official MiniMax tool?
No. It is an unofficial, fan-made estimator. MiniMax’s invoice and platform.minimax.io pricing are authoritative.
Methodology & sources: rates verified against MiniMax’s official Pricing Overview on October 24, 2026 (Asia/Shanghai) — https://platform.minimax.io/docs/pricing/overview — and cross-checked with the MiniMax M3.1 Flash launch post (https://www.minimax.io/blog/minimax-m31f). Estimates exclude taxes, reseller markups, and non-API products (M Plan, MiniMax Code). Your invoice follows metered usage.