v0.1.25 · 144 models · 22 providers

Know what you're paying for
LLM API calls

A zero-dependency Python CLI and library for comparing LLM API pricing across all major providers. No API key required.

★ Star on GitHub Install
📝
New post: How to Compare LLM API Costs Across 15 Providers with One Command
Mistral Large 3 dropped 75%. The best-value flagship is $0.50/Mtok. Here's the tool. →
bash
$ llm-prices list --provider OpenAI --sort input
Model Provider Input/Mtok Output/Mtok Context
-----------------------------------------------------------------------
gpt-4.1-nano OpenAI $ 0.1000 $ 0.4000 1047k Fastest, cheapest GPT-4.1
gpt-4o-mini OpenAI $ 0.1500 $ 0.6000 128k Small, fast, cheap
gpt-4.1-mini OpenAI $ 0.4000 $ 1.6000 1047k 1M context, cost-efficient
gpt-3.5-turbo OpenAI $ 0.5000 $ 1.5000 16k
gpt-4.1 OpenAI $ 2.0000 $ 8.0000 1047k 1M context flagship
gpt-4o OpenAI $ 2.5000 $ 10.0000 128k Latest multimodal flagship
...
$ llm-prices compare gpt-4o claude-sonnet-4-6 gemini-2.5-flash --in 5000 --out 1000
Comparison: 5,000 input tokens, 1,000 output tokens
Model Provider Input Output Total
-----------------------------------------------------------------------
gemini-2.5-flash Google $0.0008 $0.0006 $0.0014
claude-sonnet-4-6 Anthropic $0.0150 $0.0150 $0.0300 (22.0x)
gpt-4o OpenAI $0.0125 $0.0100 $0.0225 (16.5x)
$ llm-prices budget 1.00 --in 1000 --out 500
Budget: $1.0000 | Tokens per call: 1,000 in / 500 out
Model Provider Cost/call Calls
-----------------------------------------------------------
llama-3.1-8b Groq $0.000090 11,111
gemini-1.5-flash-8b Google $0.000112 8,888
gpt-4o OpenAI $0.007500 133
claude-opus-4-7 Anthropic $0.052500 19

Everything you need for LLM cost planning

🔍

list

Browse all 144 models with pricing, filterable by provider or name. Sort by input cost, output cost, or context window.

🧮

calc

Calculate exact cost for a specific number of input and output tokens on any model. JSON output for scripting.

⚖️

compare

Side-by-side cost comparison of multiple models for the same token usage. Shows relative cost multiples.

💵

budget

Given a dollar budget and typical token counts, see how many API calls each model allows. Great for cost planning.

🏆

top

Show the N cheapest models for your exact token workload. llm-prices top 5 --in 5000 --out 1000. Supports --markdown output.

📋

Markdown & CSV export

--markdown outputs a GitHub-flavored table — paste directly into your README or docs. --csv for spreadsheets.

📦

Zero dependencies

Pure Python stdlib. No numpy, no requests, no bloat. pip install and it just works.

🔌

Library API

Import calculate_cost and MODELS directly in your Python code for programmatic cost checks.

Install

# pipx (recommended — works now)
pipx install git+https://github.com/benbencodes/llm-prices

# Homebrew (macOS/Linux)
brew tap benbencodes/tap
brew install llm-prices

# From source (pip install llm-prices coming soon to PyPI)
git clone https://github.com/benbencodes/llm-prices
pip install -e llm-prices

# Then:
llm-prices list
llm-prices calc gpt-4o --in 10000 --out 2000
llm-prices compare gpt-4.1 claude-sonnet-4-6 --in 5000 --out 1000
llm-prices budget 5.00 --in 2000 --out 500

Python 3.8+. No other dependencies.

144 models across 22 providers

Pricing data is baked into the package and updated with each release.

OpenAI (GPT-4o, GPT-4.1, o1, o3, o4-mini...)
Anthropic (Claude 3, Claude 4...)
Google (Gemini 1.5, 2.0, 2.5...)
Together AI (Qwen3, Kimi K2, DeepSeek...)
Fireworks AI (DeepSeek V4 Pro, V3, Kimi...)
Cerebras (ultra-fast silicon)
SambaNova (Llama 4, RDU inference)
Amazon Bedrock (Nova Micro/Lite/Pro/Premier)
Groq / Llama 4
Perplexity AI (Sonar)
Mistral AI
Cohere
DeepSeek
xAI / Grok
AI21 Labs (Jamba — 256k context)