ChatGPT, Claude, Gemini, Grok — all four for $20.00/mo

0
View Pricing

AI Models and Credit Pricing

Home 
 
/
 AI Models and Credit Pricing

Every model available through the Bluehost AI Gateway draws from the same AI Credits wallet, but different models cost different amounts of credits to use. This article lists every supported model, grouped by provider, along with its credit cost per 1 million tokens. For an overview of how AI Credits work in general, see Understanding AI Credits.

Tip: How to read this table — all costs below are shown in credits per 1 million tokens, at the rate you're actually billed.

  • Input is the cost of tokens you send to the model.
  • Output is the cost of tokens the model generates in response.
  • Cache Write applies when a model writes context to cache for reuse.
  • Cached Input is the discounted rate for tokens read from cache instead of processed fresh.

Not every model supports caching — where a column shows a dash, that pricing tier doesn't apply to that model.

Anthropic — Claude

Anthropic Claude model credit costs, in credits per 1 million tokens
Model Input Output Cache Write Cached Input
claude-haiku-4-5 110 550 137.5 11
claude-sonnet-5 220 1,100 275 22
claude-sonnet-4-5 330 1,650 412.5 33
claude-sonnet-4-6 330 1,650 412.5 33
claude-opus-4-5 550 2,750 687.5 55
claude-opus-4-6 550 2,750 687.5 55
claude-opus-4-7 550 2,750 687.5 55
claude-opus-4-8 550 2,750 687.5 55
claude-opus-5 550 2,750 687.5 55
claude-fable-5 1,100 5,500 1,375 110
claude-opus-4-1 1,650 8,250 2,062.5 165

Google — Gemini

Google Gemini model credit costs, in credits per 1 million tokens
Model Input Output Cache Write Cached Input
gemini-3.1-flash-lite 27.5 165 — 2.75
gemini-3-flash-preview 55 330 — 5.5
gemini-3.5-flash 165 990 — 16.5
gemini-3.1-pro 220 1,320 — 22

OpenAI — GPT

OpenAI GPT model credit costs, in credits per 1 million tokens
Model Input Output Cache Write Cached Input
gpt-4o-mini 16.5 66 — 8.25
gpt-5.4-nano 22 137.5 — 2.2
gpt-5.6-luna 22 132 27.5 2.2
gpt-5.4-mini 82.5 495 — 8.25
gpt-5.2 192.5 1,540 — 19.25
gpt-5.6-terra 220 1,320 275 22
gpt-4o 275 1,100 — 137.5
gpt-5.4 275 1,650 — 27.5
gpt-5.5 550 3,300 — 55
gpt-5.6 550 3,300 687.5 55
gpt-5.6-sol 550 3,300 687.5 55

Additional Models

This group includes additional models from providers beyond Anthropic, Google, and OpenAI.

Additional model credit costs, in credits per 1 million tokens
Model Input Output Cache Write Cached Input
MiniMax-M2.5 18.15 145.2 — 3.63
DeepSeek-V4-Flash 20.9 56.1 — —
MiniMax-M3 36.3 145.2 — 7.26
Mistral-Large-3 55 165 — —
DeepSeek-V3.2 63.8 184.8 — —
Kimi-K2.6 104.5 440 — —
Kimi-K2.7-Code 104.5 440 — —
grok-4.3 137.5 275 — 22
GLM-5.2 169.4 532.4 — 16.5
DeepSeek-V4-Pro 191.4 382.8 — —
grok-4-20-reasoning 220 660 — —
Kimi-K3 363 1,815 — 36.3

Summary

Every model behind the Bluehost AI Gateway draws from the same AI Credits wallet, but credit cost per 1 million tokens varies by model and provider — lighter, faster models like claude-haiku-4-5 or gemini-3.1-flash-lite cost far less than larger models like claude-opus-4-1 or gpt-5.6-sol. Costs are broken out by input, output, cache write, and cached input where each model supports it. Check this page before choosing a default model for your app, especially if you're running high-volume workloads where the per-token cost adds up quickly.

Don't have AI Credits yet? See Purchase AI Credits for Your Server. Once you have a balance, create an API key by following How to Manage Your Bluehost API Keys.