ModelPricing.aiLLM Pricing — Cost Per Token for Every Model
How much does an LLM cost per token? This page breaks down LLM pricing for every model across OpenAI, Anthropic, and Google — from budget models to flagship reasoning engines. The table reflects the current standard rates tracked by ModelPricing.ai.
LLM Cost Per Token
Every LLM API charges per token. The table below shows the exact cost per single token for every model. For breakpoint models, the lower rate (under the context threshold) is shown. Use our LLM cost calculator for custom estimates, or see the full pricing comparison with per-million rates.
| Model | Provider | Input / token | Output / token | Input / 1M | Pricing |
|---|---|---|---|---|---|
| gpt-5-nano | OpenAI | $0.00000005 | $0.00000040 | $0.05 | Flat |
| gemini-2.0-flash-lite | Google Gemini | $0.00000007 | $0.00000030 | $0.07 | Flat |
| gemini-2.5-flash-lite | Google Gemini | $0.00000010 | $0.00000040 | $0.10 | Flat |
| gpt-4.1-nano | OpenAI | $0.00000010 | $0.00000040 | $0.10 | Flat |
| gemini-2.0-flash | Google Gemini | $0.00000015 | $0.00000060 | $0.15 | Flat |
| gpt-4o-mini | OpenAI | $0.00000015 | $0.00000060 | $0.15 | Flat |
| gpt-5.4-nano | OpenAI | $0.00000020 | $0.00000125 | $0.20 | Flat |
| gemini-3.1-flash-lite-preview | Google Gemini | $0.00000025 | $0.00000150 | $0.25 | Flat |
| claude-haiku-3 | Anthropic Claude | $0.00000025 | $0.00000125 | $0.25 | Flat |
| gpt-5-mini | OpenAI | $0.00000025 | $0.00000200 | $0.25 | Flat |
| gemini-2.5-flash | Google Gemini | $0.00000030 | $0.00000250 | $0.30 | Flat |
| gpt-4.1-mini | OpenAI | $0.00000040 | $0.00000160 | $0.40 | Flat |
| gemini-3-flash | Google Gemini | $0.00000050 | $0.00000300 | $0.50 | Flat |
| gemini-2.5-flash-preview-tts | Google Gemini | $0.00000050 | $0.00001000 | $0.50 | Flat |
| gpt-5.4-mini | OpenAI | $0.00000075 | $0.00000450 | $0.75 | Flat |
| claude-haiku-3-5 | Anthropic Claude | $0.00000080 | $0.00000400 | $0.80 | Flat |
| gemini-2.5-pro-preview-tts | Google Gemini | $0.00000100 | $0.00002000 | $1.00 | Flat |
| claude-haiku-4-5 | Anthropic Claude | $0.00000100 | $0.00000500 | $1.00 | Flat |
| o1-mini | OpenAI | $0.00000110 | $0.00000440 | $1.10 | Flat |
| o3-mini | OpenAI | $0.00000110 | $0.00000440 | $1.10 | Flat |
| o4-mini | OpenAI | $0.00000110 | $0.00000440 | $1.10 | Flat |
| gemini-2.5-pro | Google Gemini | $0.00000125 | $0.00001000 | $1.25 | Breakpoint (200K) |
| gemini-2.5-computer-use | Google Gemini | $0.00000125 | $0.00001000 | $1.25 | Breakpoint (200K) |
| gpt-5 | OpenAI | $0.00000125 | $0.00001000 | $1.25 | Flat |
| gpt-5-codex | OpenAI | $0.00000125 | $0.00001000 | $1.25 | Flat |
| gpt-5.1 | OpenAI | $0.00000125 | $0.00001000 | $1.25 | Flat |
| gpt-5.1-codex | OpenAI | $0.00000125 | $0.00001000 | $1.25 | Flat |
| gpt-5.1-codex-max | OpenAI | $0.00000125 | $0.00001000 | $1.25 | Flat |
| gemini-3.5-flash | Google Gemini | $0.00000150 | $0.00000900 | $1.50 | Flat |
| gpt-5.2 | OpenAI | $0.00000175 | $0.00001400 | $1.75 | Flat |
| gpt-5.2-codex | OpenAI | $0.00000175 | $0.00001400 | $1.75 | Flat |
| gpt-5.3-codex | OpenAI | $0.00000175 | $0.00001400 | $1.75 | Flat |
| gemini-3-pro-preview | Google Gemini | $0.00000200 | $0.00001200 | $2.00 | Breakpoint (200K) |
| gemini-3.1-pro-preview | Google Gemini | $0.00000200 | $0.00001200 | $2.00 | Breakpoint (200K) |
| gpt-4.1 | OpenAI | $0.00000200 | $0.00000800 | $2.00 | Flat |
| o3 | OpenAI | $0.00000200 | $0.00000800 | $2.00 | Flat |
| o4-mini-deep-research | OpenAI | $0.00000200 | $0.00000800 | $2.00 | Flat |
| gpt-4o | OpenAI | $0.00000250 | $0.00001000 | $2.50 | Flat |
| gpt-5.4 | OpenAI | $0.00000250 | $0.00001500 | $2.50 | Breakpoint (272K) |
| claude-3-7-sonnet | Anthropic Claude | $0.00000300 | $0.00001500 | $3.00 | Flat |
| claude-sonnet-4-0 | Anthropic Claude | $0.00000300 | $0.00001500 | $3.00 | Flat |
| claude-sonnet-4-5 | Anthropic Claude | $0.00000300 | $0.00001500 | $3.00 | Flat |
| claude-sonnet-4-6 | Anthropic Claude | $0.00000300 | $0.00001500 | $3.00 | Flat |
| claude-opus-4-5 | Anthropic Claude | $0.00000500 | $0.00002500 | $5.00 | Flat |
| claude-opus-4-6 | Anthropic Claude | $0.00000500 | $0.00002500 | $5.00 | Flat |
| claude-opus-4-7 | Anthropic Claude | $0.00000500 | $0.00002500 | $5.00 | Flat |
| claude-opus-4-8 | Anthropic Claude | $0.00000500 | $0.00002500 | $5.00 | Flat |
| gpt-5.5 | OpenAI | $0.00000500 | $0.00003000 | $5.00 | Breakpoint (272K) |
| claude-fable-5 | Anthropic Claude | $0.00001000 | $0.00005000 | $10.00 | Flat |
| claude-mythos-5 | Anthropic Claude | $0.00001000 | $0.00005000 | $10.00 | Flat |
| o3-deep-research | OpenAI | $0.00001000 | $0.00004000 | $10.00 | Flat |
| claude-opus-3 | Anthropic Claude | $0.00001500 | $0.00007500 | $15.00 | Flat |
| claude-opus-4-0 | Anthropic Claude | $0.00001500 | $0.00007500 | $15.00 | Flat |
| claude-opus-4-1 | Anthropic Claude | $0.00001500 | $0.00007500 | $15.00 | Flat |
| gpt-5-pro | OpenAI | $0.00001500 | $0.0001200 | $15.00 | Flat |
| o1 | OpenAI | $0.00001500 | $0.00006000 | $15.00 | Flat |
| o3-pro | OpenAI | $0.00002000 | $0.00008000 | $20.00 | Flat |
| gpt-5.2-pro | OpenAI | $0.00002100 | $0.0001680 | $21.00 | Flat |
| gpt-5.4-pro | OpenAI | $0.00003000 | $0.0001800 | $30.00 | Breakpoint (272K) |
| gpt-5.5-pro | OpenAI | $0.00003000 | $0.0001800 | $30.00 | Flat |
| o1-pro | OpenAI | $0.0001500 | $0.0006000 | $150.00 | Flat |
How LLM Token Pricing Works
LLM APIs charge per token — a unit of text roughly equal to 4 characters or 0.75 words in English. Every API call has two cost components:
- Input tokens (your prompt) — processed in parallel, relatively cheap.
- Output tokens (the model's response) — generated sequentially, typically 2-8x more expensive than input.
Some models use breakpoint pricing, where rates increase above a context length threshold (typically 200K tokens). This means a long conversation or document analysis costs more per token than a short prompt. Models with breakpoint pricing are marked in the table above.
Multimodal models (image and audio) have separate rates for each modality. Image output tokens can cost 10-60x more than text output on the same model. See individual provider pages for multimodal pricing details.
Provider-specific modifiers such as batch discounts, priority/fast modes, and data residency uplifts can change the final bill. This table focuses on the standard per-token rates used by the estimator.
LLM Pricing by Provider
Frequently Asked Questions
How much does an LLM cost per token?
In the current table, standard text input costs start at $0.00000005 per token for GPT-5 Nano and can reach $0.000150 per token for premium reasoning models. Most mid-range models sit around $0.000001-$0.000003 per input token before cache, batch, priority, or residency modifiers.
What is the cheapest LLM?
The cheapest standard-rate models in the table are GPT-5 Nano ($0.05/1M input), Gemini 2.0 Flash Lite ($0.075/1M input), and the low-cost nano/flash tiers around $0.10/1M input. They are best suited to high-volume, low-complexity work.
Why are output tokens more expensive than input tokens?
Output tokens require the model to generate text autoregressively — each token depends on all previous tokens. This is more computationally expensive than processing input tokens in parallel. Output tokens typically cost 2-8x more than input tokens.
How do I calculate my LLM API cost?
Multiply your input token count by the input rate, and your output token count by the output rate. For example, 1,000 input tokens + 500 output tokens on GPT-5 ($1.25/1M input, $10/1M output) costs: (1000 × $0.00000125) + (500 × $0.00001) = $0.00625. Use our LLM cost calculator for instant estimates.
Automate Your LLM Cost Estimation
Get programmatic access to real-time LLM pricing with our API. Or try the cost calculator and pricing comparison for quick estimates.
Get Started Free