LLM API price comparison
Per-1M-token prices for 80+ chat models from Anthropic, OpenAI, Google, xAI, DeepSeek, Qwen, Moonshot AI, Z.ai, MiniMax and more: the developer's list price next to the UplinkSwitch price. It reads the public catalog at api.uplinkswitch.com/v1/catalog on every load. Daily snapshots are in the LLM API Pricing dataset.
Prices by family: Claude · GPT · Gemini · Grok · DeepSeek · Kimi · GLM · Qwen · Compare two models
| Model | Model id | Developer | List in | List cached | List out | Uplink in | Uplink cached | Uplink out | Under list (out) | List / month | Uplink / month | Weights |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Loading… | ||||||||||||
| Model | Model id | Developer | Kind | Variant | Per | List $ | Uplink $ | Under list |
|---|---|---|---|---|---|---|---|---|
| Loading… | ||||||||
Use these prices from your code
UplinkSwitch speaks the OpenAI and Anthropic API formats. Point your SDK's base URL at https://api.uplinkswitch.com/v1 and use the model id from the table.
from openai import OpenAI
client = OpenAI(base_url="https://api.uplinkswitch.com/v1", api_key="sk-uplink-live-...")
reply = client.chat.completions.create(
model="claude-opus-5-5",
messages=[{"role": "user", "content": "hi"}],
)
print(reply.choices[0].message.content)Setup guides: OpenAI, Anthropic and Gemini SDKs · Claude Code · Cursor · Cline · Roo Code · Kilo Code · OpenCode · Veo 3.1 · Nano Banana Pro
Read more
- LLM API pricing comparison, October 2026
- Claude API pricing · GPT API pricing · Gemini API pricing · Grok API pricing
- DeepSeek API pricing · Kimi API pricing · GLM API pricing · Qwen API pricing
- What Claude Code costs on API billing · Using Cursor with your own API key: what it costs
- Kimi, GLM and Qwen coding plans vs pay-per-token API · The hidden prompt tokens in cheap LLM APIs
- Open weights for every open model in the table: LLM API pricing: open-weight models
Questions
Where do the list prices come from?
They're each developer's published standard-tier price per million tokens, stored in the catalog next to the gateway price. Batch, priority and long-context tiers aren't shown. Models without a published price have empty list columns.
How current is this table?
It reads the live catalog when the page loads, the same feed the gateway bills from. The timestamp under the table shows when the catalog was generated.
Why is cached input sometimes above list?
On a few models the gateway's cached-input price is higher than the developer's, even though input and output are lower. If your workload reads mostly from cache, raise the cache share in the calculator and compare the monthly columns.
Do the prices include volume discounts?
No. Monthly volume takes 3% off from 100M tokens, 6% from 1B and 10% from 10B. The table shows base prices.
Does it work with Claude Code, Cursor or Cline?
Yes. Each tool takes a custom base URL and key. See the setup guides.