NNyquest docs

Plans and Pricing

Nyquest doesn't have monthly subscription tiers. Pricing is pay-as-you-go from a wallet, with one upgrade event (free β†’ Pro) triggered by your first funding.

The two tiers

FreePro
Cost to upgradeβ€”First successful funding (minimum $5 through Checkout)
Chat rate limit20/min120/min
Auth rate limitShared IP limits: 5/min; sign-in 10/hour; signup 3/hourSame
Live support accessNoYes (via /help/live-support)
Artifact storage / file10MB100MB
Total artifact storageLimitedGenerous (100MB+)
BYOKβœ…βœ…
Agent ModeAvailable; paid planner calls require wallet fundsAvailable; paid planner calls require wallet funds
Multi-Model SplicerβŒβœ…
Memory layerβœ…βœ…
Projectsβœ…βœ…
Model catalogBrowse all; use available free routes or BYOKFunded access, subject to availability

One feature is Pro-and-up: the Multi-Model Splicer. On the free tier a splice is rejected with 402 upgrade_required. Everything else is shared, and the remaining differences are operational: Pro gets higher rate limits and live support.

What costs what

Platform-hosted chat (we use our keys)

Pricing matches the provider's published per-token rates with no markup on top.

What makes Nyquest different is Split the Diff. Your prompt is compressed before it reaches the model, and the tokens compression removes β€” priced at that model's input rate β€” are the diff. On eligible managed chat requests, it is split 50/50: your half is credited straight back to your wallet in the same transaction (it appears as Split the Diff in your transaction history), the other half is Nyquest's fee. The green ⚑ pill in the top bar keeps a running total of what you've been credited, and the Savings page (/savings) has the full ledger β€” per model and per request.

Two honest footnotes. A credit never takes a request below what the provider charged Nyquest for it, so when memory or web-grounding context you weren't billed for makes a request cost more upstream, the credit is held at that floor. And compression is skipped where it doesn't help β€” media generation, free routes, and a handful of models where evaluation showed it hurt quality. BYOK requests never touch your wallet; those savings are entirely yours at your provider.

Use the live pricing catalog and model picker for a specific model’s current input/output prices. A tier name is not a price: models within the same tier can have substantially different rates.

BYOK chat

Free on the platform. Your provider charges you directly via your API key β€” Nyquest is just routing the request. We don't add a fee.

Agent Mode

Same per-token pricing as chat, but multiple LLM calls happen per agent run (one per step and, when needed, a final synthesis call). Each paid planner call reserves a conservative amount before dispatch and settles its actual user charge afterward. A run can stop before its next call if funds or budget are insufficient. See Costs and Limits.

Image generation

Use the available models in the image picker or live catalog. Images are billed at the provider's actual charge, without a Nyquest surcharge. Google retired its direct Imagen 4 models; those choices are no longer offered. Existing generated images remain in your gallery.

If the provider cost is missing, the result shows Cost pending until it can be reconciled; no estimated debit is applied. See Media Studio.

TTS (text-to-speech)

Gemini TTS also runs directly on Nyquest's own provider access. It is billed at the selected model's published input text and output audio token rates, using the provider's reported token counts. Generate it in Media Studio.

Live support

Free for Pro users. Goes through the live operator (Mike) via Discord-bridged thread.

What about volume discounts

Currently: none. Pricing is flat per-token regardless of volume.

If you're at the level where this matters ($1,000+/month), talk to a real human β€” we can discuss enterprise arrangements case-by-case.

What's NOT charged

  • Account creation (free)
  • Sign-in / sign-out / OAuth (free)
  • Browsing the help section (free)
  • Reading past conversations (free)
  • Browsing your artifact vault (free)
  • Switching chassis (free)
  • Creating, editing, deleting projects (free)
  • Memory operations (recall is free; building memories happens during chat which IS billed for the underlying chat call but not for the memory layer specifically)

When billing happens

  • Real-time for chat: each message decrements the wallet at the moment the response completes
  • Per call for agent runs: a planner call reserves funds before dispatch and settles after its response
  • At completion for image generation / TTS when provider cost is known; missing cost remains pending reconciliation
  • No monthly bill β€” there's nothing recurring to bill for

Where to next