Claude Haiku tier
The displayed $0.10 / $0.50 rate is for prompts up to 100K tokens. The Claude page shows $0.50 input and $2.50 output above 100K, so split or reprice those requests.
Read the Haiku tier guide →MULTI-PROVIDER API · LAST VERIFIED: 2026-10-31
LLM API Cost
Comparison
Put one monthly workload beside published OpenAI, Claude and Gemini token rates. Change the input/output mix and the comparison updates in your browser. This is an independent unofficial estimate; prices can change.
01 / SAME WORKLOAD
Compare the cost of one workload.
Start with 10M input + 2M output tokens, or enter your own monthly totals. No API key or usage data is sent anywhere.
Presets
Formula: (input tokens ÷ 1M × input rate) + (output tokens ÷ 1M × output rate).
Lowest estimate in this table$2.00GPT-6 Luna · 10M input + 2M output
02 / RATE CARD + ESTIMATE
Every row uses the same tokens.
The initial HTML includes the default 10M input + 2M output result. Change the fields above to recalculate every estimate locally.
| Model | Provider / route | Input / 1M | Output / 1M | Estimated cost | Source |
|---|---|---|---|---|---|
| GPT-6 Astra gpt-6-astra | OpenAI | $10.00 | $50.00 | $200.00 | OpenAI card |
| GPT-6.1 Sol gpt-6.1-sol | OpenAI | $2.00 | $10.00 | $40.00 | OpenAI card |
| GPT-6.1 Terra gpt-6.1-terra | OpenAI | $1.00 | $5.00 | $20.00 | OpenAI card |
| GPT-6 Luna gpt-6-luna | OpenAI | $0.10 | $0.50 | $2.00 | OpenAI card |
| GPT-5.6 Sol gpt-5.6-sol | OpenAI | $4.00 | $20.00 | $80.00 | OpenAI card |
| Claude Haiku 5.5 claude-haiku-5-5 | Claude Platform · ≤100K prompt | $0.10 | $0.50 | $2.00 | Claude card |
| Claude Sonnet 5.5 claude-sonnet-5-5 | Claude Platform | $2.00 | $10.00 | $40.00 | Claude card |
| Claude Opus 5.5 claude-opus-5-5 | Claude Platform | $4.00 | $20.00 | $80.00 | Claude card |
| Gemini 4 Argon gemini-4-argon | Google Gemini API · introductory | $2.00 | $10.00 | $40.00 | Gemini card |
The same token count can produce different product outcomes. Include retries, extra turns, prompt caching, tool calls and quality requirements when you budget a real application.
03 / READ THE RESULT
Token price is a useful baseline, not an invoice.
These rows are comparable only within the assumptions shown.
Gemini date window
Gemini 4 Argon is shown at the introductory $2 input and $10 output rate through November 30, 2026. The existing Gemini page records the announced later $4 / $20 rate; recheck before December budgets.
Read the Gemini guide →OpenAI modifiers
The OpenAI rows use standard short-context list rates. Cached input, Batch or Flex modes, long context, tools and other routes can change the total.
OpenAI calculator →04 / FAQ
LLM API cost comparison questions
How does this LLM API cost comparison work?
Enter monthly input and output token totals. Each row calculates input tokens divided by one million times the input rate, plus output tokens divided by one million times the output rate.
Which prices are included here?
This page only reuses prices already shown on this site: five OpenAI models from the OpenAI calculator, three Claude models from the Claude pricing page, and Gemini 4 Argon's introductory rates from the Gemini pricing page.
Does the result show my exact invoice?
No. It is an independent unofficial token estimate. It excludes cache charges, batch or priority adjustments, tools, taxes, hosting and provider-specific fees. Verify the exact official rate card before deployment.
Can I compare a prompt over 100K tokens with the Haiku row?
Not at the displayed short-tier rate. The Haiku row is for prompts up to 100K tokens. For prompts over 100K, use the higher rates shown in the Haiku pricing guide.
Where can I see the wider LLM rate card?
Visit the LLM API pricing comparison for the broader site guide, then use this page when you need one workload calculated across multiple providers.
Need a provider-specific estimate?
Open the detailed OpenAI calculator, Claude rate card, or Gemini pricing guide.