OpenAI-Compatible LLM API Router
Keep the OpenAI SDK. Change base_url once and call hundreds of models through one API. Create a key and make the first request without a card.
Keep the SDK. Change the base URL.
Use the OpenAI client you already have. Change base_url, create one key, and call hundreds of models through the same interface.
- ✓ Install the standard
openaipackage - ✓ Point it at
https://api.trustedrouter.com/v1 - ✓ Start with
trustedrouter/cheapor choose any model ID
No card required. Your existing OpenAI SDK keeps working.
import os
from openai import OpenAI
client = OpenAI(
api_key=os.environ["TRUSTEDROUTER_API_KEY"],
base_url="https://api.trustedrouter.com/v1",
)
response = client.chat.completions.create(
model="trustedrouter/cheap",
messages=[{"role": "user", "content": "Reply with PONG"}],
)
print(response.choices[0].message.content)
Measured providers
Use live TTFT, throughput, and uptime data instead of vendor claims.
Open leaderboardPolicy-aware routing
Compare ZDR, confidential compute, provider E2EE, and upstream policy sources.
Open providersVerify the gateway
Check the running image, source commit, digest, and attestation path.
Open trustCurrent routes, prices, privacy, and measured performance.
Catalog facts come from the routes currently configured in TrustedRouter. Performance uses the same cached metadata snapshot as the public leaderboard. Prompts and outputs are not part of these measurements.
| Model | Providers | Context | Input | Output | Privacy | Measured route |
|---|---|---|---|---|---|---|
Anthropic: Claude Opus 4.8anthropic/claude-opus-4.8 |
3 routes | 1,000,000 | $5.275/1M | $26.375/1M | varies 5 cited scores | 1735 ms TTFT anthropic · 100.00% available · n=4 |
OpenAI: GPT-5.5openai/gpt-5.5 |
4 routes | 1,050,000 | $5.275/1M | $31.65/1M | ZDR 3 cited scores | Warming up |
Google: Gemini 3.5 Flashgoogle/gemini-3.5-flash |
7 routes | 1,048,576 | $1.5825/1M | $9.495/1M | ZDR | 1372 ms TTFT google-vertex · 58 tok/s · 100.00% available · n=8 |
MoonshotAI: Kimi K2.7 Codemoonshotai/kimi-k2.7-code |
19 routes | 262,144 | $0.70685/1M to $1.00225/1M | $3.587/1M to $4.22/1M | ZDR 5 cited scores | 2918 ms TTFT kimi · 100.00% available · n=23 |
Z.ai: GLM 5.2z-ai/glm-5.2 |
44 routes | 1,048,576 | $0.7174/1M to $2.434729/1M | $1.5825/1M to $7.030281/1M | E2EE 4 cited scores | 4944 ms TTFT baseten · 100.00% available · n=71 |
MiniMax: MiniMax M3minimax/minimax-m3 |
21 routes | 524,288 | $0.24265/1M to $0.633/1M | $1.0128/1M to $2.532/1M | ZDR 4 cited scores | 1599 ms TTFT together · 137 tok/s · 100.00% available · n=71 |
Browse every modelReview provider policiesOpen the full leaderboardSnapshot 2026-09-16T06:07:37.348Z
Questions
Can I keep using the OpenAI SDK?
Yes. Set base_url to the TrustedRouter API and keep the usual chat completions or Responses API calls.
Can I still choose exact providers?
Yes. Use explicit model IDs, provider filters, or TrustedRouter aliases such as trustedrouter/auto and trustedrouter/zdr.