Claud next to Claude, GPT and Gemini
Our published per-token prices beside the vendors' own list prices, on the same workload, with the date each figure was checked and a link to where it came from. Across the models below, Claud is typically 70% cheaper.
Claud is an independent service. Claud models are our own; they are not Claude, GPT or Gemini, and we are not affiliated with Anthropic, OpenAI or Google. Vendor names are used only to identify what is being compared.
Claud vs Anthropic
Anthropic list prices as published on 2025-10-01 · source
| Anthropic model | Claud model | Input | Cached input | Output | Reference workload |
|---|---|---|---|---|---|
Claude Opus 4.1 200K context · Standard API tier; batch and long-context surcharges not included. | Claud Opus claud-opus · 256K context | $15.00 Claud $3.00 | $1.500 Claud $0.150 | $75.00 Claud $9.00 | 85% less $4.39 vs $29.70 |
Claude Sonnet 4.5 200K context · Prompts over 200K tokens are billed at a higher long-context rate. | Claud 5.1 claud-5.1 · 256K context | $3.00 Claud $0.75 | $0.300 Claud $0.075 | $15.00 Claud $3.00 | 78% less $1.29 vs $5.94 |
Claude Sonnet 4.5 (extended thinking) 200K context · Thinking tokens are billed as output at the same rate. | Claud Reason claud-reason · 256K context | $3.00 Claud $0.90 | $0.300 Claud $0.090 | $15.00 Claud $3.60 | 74% less $1.55 vs $5.94 |
Claude Haiku 4.5 200K context | Claud Flash claud-flash · 128K context | $1.00 Claud $0.22 | $0.100 Claud $0.022 | $5.00 Claud $1.75 | 70% less $0.59 vs $1.98 |
Claud vs OpenAI
OpenAI list prices as published on 2025-10-01 · source
| OpenAI model | Claud model | Input | Cached input | Output | Reference workload |
|---|---|---|---|---|---|
GPT-5 pro 400K context · Responses API only; no cached-input rate published. | Claud Opus claud-opus · 256K context | $15.00 Claud $3.00 | — Claud $0.150 | $120.00 Claud $9.00 | 90% less $4.39 vs $45.00 |
GPT-5 400K context · Reasoning tokens are billed as output. | Claud 5.1 claud-5.1 · 256K context | $1.25 Claud $0.75 | $0.125 Claud $0.075 | $10.00 Claud $3.00 | 62% less $1.29 vs $3.41 |
o3 200K context · Reasoning tokens are billed as output. | Claud Reason claud-reason · 256K context | $2.00 Claud $0.90 | $0.500 Claud $0.090 | $8.00 Claud $3.60 | 56% less $1.55 vs $3.55 |
GPT-5 mini 400K context | Claud Flash claud-flash · 128K context | $0.25 Claud $0.22 | $0.025 Claud $0.022 | $2.00 Claud $1.75 | 12% less $0.59 vs $0.68 |
Claud vs Google
Google list prices as published on 2025-10-01 · source
| Google model | Claud model | Input | Cached input | Output | Reference workload |
|---|---|---|---|---|---|
Gemini 2.5 Pro 1049K context · Rates for prompts up to 200K tokens; larger prompts cost more. Paid tier. | Claud 5.1 claud-5.1 · 256K context | $1.25 Claud $0.75 | $0.310 Claud $0.075 | $10.00 Claud $3.00 | 63% less $1.29 vs $3.46 |
Gemini 2.5 Flash 1049K context · Paid tier; output rate includes thinking tokens. | Claud Flash claud-flash · 128K context | $0.30 Claud $0.22 | $0.030 Claud $0.022 | $2.50 Claud $1.75 | 29% less $0.59 vs $0.84 |
Reference workload: 1.00M input tokens with a 30% cache hit rate and 250.0K output tokens. Grey figures under each price are the Claud rate. Where a vendor is cheaper we say so.
Plans: cheaper again
The tables above use our standard pay-as-you-go rate of $0.000002 per Claud token. A plan buys the monthly allowance below that rate, and because every model is billed in the same tokens, every price on this page drops by the same factor. Prepay a quarter for 5% off or a year for 12% off and it drops further. On the flagship pairing (Claud Opus vs GPT-5 pro) that turns a 90% saving into 93% on the largest plan.
| Plan | Per month | Tokens / month | Effective token price | vs standard rate | Claud Opus vs GPT-5 pro | Claud Flash vs GPT-5 mini |
|---|---|---|---|---|---|---|
Starter billed monthly · quarterly −5% · annual −12% | $20 | 12.00M | $0.0000017 | 17% less | 92% less $3.66 vs $45.00 | 27% less $0.49 vs $0.68 |
Pro billed monthly · quarterly −5% · annual −12% | $60 | 39.00M | $0.0000015 | 23% less | 92% less $3.38 vs $45.00 | 33% less $0.46 vs $0.68 |
Power billed monthly · quarterly −5% · annual −12% | $200 | 140.00M | $0.0000014 | 29% less | 93% less $3.13 vs $45.00 | 37% less $0.42 vs $0.68 |
Business billed monthly · quarterly −5% · annual −12% | $700 | 525.00M | $0.0000013 | 33% less | 93% less $2.93 vs $45.00 | 42% less $0.39 vs $0.68 |
Plan columns apply the plan's effective token price to the reference workload above. Unused allowance rolls within the month only; purchased token packages never expire and can be combined with any plan.
Your workload
Put in your own monthly volume and see the bill side by side.
Share of input served from prompt cache.
| Model | Monthly cost | vs Claud |
|---|---|---|
Claud Opus claud-opus | $411 | — |
Claude Opus 4.1 Anthropic list price, checked 2025-10-01 | $2,835 | Claud is 86% cheaper save $2,424 / month |
GPT-5 pro OpenAI list price, checked 2025-10-01 | $4,500 | Claud is 91% cheaper save $4,089 / month |
Estimate based on published per-token list prices. Cached input is billed at each vendor's cache-read rate when one is published, otherwise at the full input rate. Vendor batch discounts, long-context surcharges, tier minimums and taxes are not included.
Beyond the per-token price
Things that change what you actually pay and how you can pay it.
Pay in crypto, no KYC
BTC, ETH, BNB, USDT and USDC (ERC-20, TRC-20, BEP-20) and more across Ethereum, Tron, BSC, Base, Arbitrum, Optimism, Polygon and Avalanche, credited automatically when the transaction confirms. No card, no identity checks, no billing account to be approved.
One balance, every model
Everything is billed in Claud tokens worth $0.000002 each, so you can switch models without opening a new account or moving credit around.
No minimums, no expiry
Purchased tokens never expire and start at $15. There are no monthly minimums, tier gates or prepaid commitments to reach a price.
Drop-in compatible API
OpenAI Chat Completions and Anthropic Messages request formats, with streaming and tool use. Change the base URL and key; your existing client code keeps working.
Chat and API together
Web chat, the playground and the API draw from one balance. No separate seat licence for the chat app and no separate billing account for the API.
Free tokens for following along
50.0K tokens for confirming the newsletter, and more for joining us on Telegram, YouTube and X.
Earn 20% by referring
Every account has a referral link. Purchases by people you refer pay you a commission, settled automatically in stablecoins or tokens.
How to read this page
- Vendor figures are the standard pay-as-you-go list prices from each vendor's public pricing page on the date shown. They exclude batch discounts, long-context surcharges, enterprise contracts and taxes, and they may have changed since we last checked.
- Claud figures are the prices currently published on our pricing page and are exactly what your account is billed.
- Models are paired by role (flagship, everyday, reasoning, fast), not by identity. Capabilities, quality and speed differ between vendors; price is only one input to the decision. The smallest token package lets you test every model before committing to a plan.
- All Claud purchases are paid in crypto and are final. There is no free tier: what keeps prices low is that we do not carry card fees, chargebacks, KYC or unpaid trial usage, and we pass that on.