§ 3 · Pricingmodel catalog · micro-PLMB per 1M tokens
NOTES
Per-model pricing, in micro-PLMB.
One PLMB = 1,000,000 micro-PLMB. Every completion debits (prompt_tokens * prompt_price + completion_tokens * completion_price) / 1M micro-PLMB from your credit balance in the same transaction as the receipt. Pricing matches upstream provider list prices plus a 0% margin during the preview period. Hub-hosted models are billed at a flat default tier until metered rates land.
| Provider · model id | Prompt / 1M | Completion / 1M |
|---|---|---|
| openai/gpt-4o GPT-4o | 2,500,000 μ | 10,000,000 μ |
| openai/gpt-4o-mini GPT-4o mini | 150,000 μ | 600,000 μ |
| openai/o1 o1 | 15,000,000 μ | 60,000,000 μ |
| anthropic/claude-sonnet-4.5 Claude Sonnet 4.5 | 3,000,000 μ | 15,000,000 μ |
| anthropic/claude-haiku-4.5 Claude Haiku 4.5 | 800,000 μ | 4,000,000 μ |
| anthropic/claude-opus-4.7 Claude Opus 4.7 | 15,000,000 μ | 75,000,000 μ |
| google/gemini-2.5-pro Gemini 2.5 Pro | 1,250,000 μ | 5,000,000 μ |
| google/gemini-2.5-flash Gemini 2.5 Flash | 75,000 μ | 300,000 μ |
| xai/grok-4 Grok 4 | 3,000,000 μ | 15,000,000 μ |
| hub/default Hub model (default tier) | 500,000 μ | 500,000 μ |
| mock/echo-v1 Mock echo (no API key required) | 1,000 μ | 1,000 μ |
↳ mock/echo-v1 is a local dev tier that needs no upstream API key.
↳ hub/default applies to any user-uploaded model via Hub.
↳ Credit top-ups go through Settlement.deposit() on Base Sepolia during preview.