Lower your inference bill.
Compare published rates for the models you already use, including input, output, and available cache pricing.
BUILT FOR FOUNDERS & GROWING TEAMS
Access frontier models,
up to 80% cheaper
Use leading AI models through one OpenAI-compatible API.
Keep your existing tools and spend less on inference.
Published token ratesOpenAI-compatible APIFor individuals and businesses
PRICING · PER 1M TOKENS
Every model, a fraction of the price
Search the catalogue, compare published rates, and explore cache pricing for your workload.
UnbarOpenRouter
Model maker
Sort models
40 of 40 models
| Model | Input | Cached input | Output | Savings | Details |
|---|---|---|---|---|---|
 Claude Fable 5Anthropic | $10.00$2.0080% less than OpenRouter | $1.00$0.2080% less than OpenRouter | $50.00$10.0080% less than OpenRouter | 80% less | |
Claude Haiku 4.5Anthropic | $1.10$0.2280% less than OpenRouter | $0.11$0.02280% less than OpenRouter | $5.50$1.1080% less than OpenRouter | 80% less | |
Claude Opus 4.6Anthropic | $5.00$1.0080% less than OpenRouter | $0.50$0.1080% less than OpenRouter | $25.00$5.0080% less than OpenRouter | 80% less | |
Claude Opus 4.7Anthropic | $5.00$1.0080% less than OpenRouter | $0.50$0.1080% less than OpenRouter | $25.00$5.0080% less than OpenRouter | 80% less | |
Claude Opus 4.8Anthropic | $5.00$1.0080% less than OpenRouter | $0.50$0.1080% less than OpenRouter | $25.00$5.0080% less than OpenRouter | 80% less | |
Claude Opus 5Anthropic | $5.00$1.0080% less than OpenRouter | $0.50$0.1080% less than OpenRouter | $25.00$5.0080% less than OpenRouter | 80% less | |
Claude Sonnet 4.6Anthropic | $3.00$0.6080% less than OpenRouter | $0.30$0.0680% less than OpenRouter | $15.00$3.0080% less than OpenRouter | 80% less | |
Claude Sonnet 5Anthropic | $2.00$0.4080% less than OpenRouter | $0.20$0.0480% less than OpenRouter | $10.00$2.0080% less than OpenRouter | 80% less | |
 DeepSeek V4 FlashDeepSeek | $0.14$0.02880% less than OpenRouter | $0.03$0.00680% less than OpenRouter | $0.28$0.05680% less than OpenRouter | 80% less | |
DeepSeek V4 ProDeepSeek | $1.74$0.34880% less than OpenRouter | $0.14$0.02880% less than OpenRouter | $3.48$0.69680% less than OpenRouter | 80% less | |
 Gemini 2.5 FlashGoogle | $0.30$0.0680% less than OpenRouter | $0.03$0.00680% less than OpenRouter | $2.50$0.5080% less than OpenRouter | 80% less | |
Gemini 2.5 ProGoogle | $1.25$0.2580% less than OpenRouter | $0.12$0.02480% less than OpenRouter | $10.00$2.0080% less than OpenRouter | 80% less |
Page 1 of 4
USD per million tokens. Crossed-out rates are OpenRouter list prices. Savings show the lower of input and output savings; individual rates can differ. Expand a row for all published rates.
Keep your existing SDK.
Use your OpenAI-compatible client with a new base URL and API key. Choose a model for each request.
Capacity for your workload.
Tell us which models you use and how much traffic you expect, so we can plan capacity for your application.
OpenAI-compatible requests
Choose your model per request
Keep your existing workflow
A FAMILIAR STARTING POINT
Change two values.
Keep your stack.
After onboarding, update your base URL and API key in your OpenAI-compatible client. Check the model and features your application uses before switching traffic.
Read the integration docsclient.py
from openai import OpenAI
import os
client = OpenAI(
base_url="https://api.ferrix.sh/v1",
api_key=os.environ["FERRIX_API_KEY"],
)Create your API key after onboarding.
Pricing, privacy, and support.
Know what to expect.
Before you move a workload, know how pricing works, what happens to your data, and what your team needs.
Trust
Check models, rates, and usage.
Public rates, usage records, and practical checks before you move production traffic.
See what you can verifyPrivacy
Zero prompt retention.
Unbar never stores your API prompts or completions. Usage and billing records stay separate from conversation content.
Read the Privacy PolicyFor teams
Billing and onboarding for teams.
Tell us your model needs and billing requirements. We confirm the options we can support before onboarding.
Explore team optionsLet’s talk about your AI spend
Change one line tonight. Watch the bill drop.
Tell us which models you use and roughly what you spend each month. An estimate is enough.
We’ll review your request and email you when we can onboard your workload.
- 01Share your models and monthly spend
- 02We review model fit and capacity
- 03Receive your invitation by email