LumeAPI
Get API key

OpenRouter alternative

OpenRouter Alternative — 5.5% Credit Fee vs Catalog Token Rates

OpenRouter passes provider list prices through and charges 5.5% (minimum $0.80) when you buy credits by card — that is a funding fee, not a token markup. LumeAPI debits a USD wallet at published catalog rates with no credit-purchase fee.

Last verified 2026-08-14 against openrouter.ai/docs/faq. Stay on OpenRouter for 100+ models, auto-fallback, or ZDR. This page is for teams whose production set is already GPT, Claude, and Gemini.

  • One API key
  • Dozens of models
  • Text, image & video
  • Up to 70% off
Create API keyReview the SDK migrationMigration checklist

LumeAPI catalog vs official reference (flagship models)

Official reference vs LumeAPI catalog rates. Pricing unit: per 1M input / output tokens. Last updated: 2026-07-22. Source: provider list price.

ModelOfficial (in / out)LumeAPI (in / out)Savings
GPT-5.6 Terragpt-5.6-terra$2.50 / $15.00$0.75 / $4.5070% offUse catalog id
Claude Sonnet 4.6claude-sonnet-4-6$3.00 / $15.00$1.50 / $7.5050% offUse catalog id
Gemini 3.1 Progemini-3.1-pro-preview$2.00 / $12.00$1.00 / $6.0050% offUse catalog id
Gemini 3.5 Flashgemini-3.5-flash$1.50 / $9.00$0.75 / $4.5050% offUse catalog id

Should you switch from OpenRouter?

Last verified: 2026-08-15

The useful answer is conditional. LumeAPI is a focused production alternative, not a drop-in replacement for every OpenRouter feature.

Your requirementBetter defaultReason
A stable set of GPT, Claude, and Gemini modelsLumeAPISmall catalog, explicit ids, and published catalog billing
Hundreds of models or weekly experimentsOpenRouterThe broader catalog is the product advantage
Provider or model fallback managed by the gatewayOpenRouterLumeAPI fails closed; your application owns fallback
ZDR-aware provider routingOpenRouterOpenRouter exposes provider-level privacy routing controls
One OpenAI-shaped client for mapped production modelsLumeAPIChange the key, base URL, and mapped model id

Do not move an unmapped route. Keep both backends during the measurement window.

Separate the funding fee from token price

Last verified: 2026-08-15

OpenRouter documents a 5.5% fee with a $0.80 minimum when credits are purchased and says underlying provider prices are passed through without markup.

Credits purchasedDocumented purchase feeWhat this does not prove
$10$0.80 minimumThat another gateway is cheaper for your model mix
$100$5.50That retries, caching, or output length are equal
$800$44.00That every OpenRouter route exists on LumeAPI
$2,000$110.00That a migration will improve task success rate

Compare a 7–14 day export from both systems. The decision metric is all-in cost per successful task, with the same prompts and retry policy.

Map model ids before sending shadow traffic

Last verified: 2026-08-15

OpenRouter uses provider-prefixed slugs. LumeAPI uses catalog ids. Copy exact strings from /models instead of guessing.

OpenRouter-style slugLumeAPI catalog idAction
openai/gpt-5.6-terragpt-5.6-terraRemove the provider prefix
anthropic/claude-sonnet-4.6claude-sonnet-4-6Use the catalog hyphenated version
google/gemini-3.5-flashgemini-3.5-flashUse the exact catalog id
Any slug absent from /modelsNo mappingKeep that route on OpenRouter

A reversible 14-day migration

Last verified: 2026-08-15

This sequence answers the user’s real question: how to test a cheaper alternative without creating an outage.

  • Inventory the OpenRouter slugs used in production, not every slug in configuration files
  • Map only models present in the live LumeAPI catalog and keep unmapped routes unchanged
  • Add a backend feature flag so each request can use OpenRouter or LumeAPI
  • Shadow 5–10% of matched traffic with identical prompts, timeouts, and retry limits
  • Compare task success, p95 latency, retries, output tokens, and all-in spend for 7–14 days
  • Increase traffic model by model; retain instant rollback until the observation window closes

Fair comparison, not a hit piece

OpenRouter earned its position as the default “one interface, many models” option. It is strong when you need broad model coverage, flexible routing, and quick access to niche endpoints. LumeAPI is narrower on purpose: mainstream GPT, Claude, and Gemini at lower published catalog rates with exact model ids in Usage logs.

Teams switch when credit funding fees and pass-through list prices compound on agents with long context. If you only call five flagship models in production, paying aggregator overhead on every hop is hard to justify.

Run OpenRouter and LumeAPI in parallel during migration. Compare cost per completed task, not single-call latency alone. Keep OpenRouter as fallback for models that are not in our catalog yet.

Migrate in minutes

Three values change: API key, base URL, model id. Everything else stays the same.

Before (OpenRouter)

python
from openai import OpenAI
client = OpenAI(api_key="YOUR_OPENROUTER_KEY", base_url="https://openrouter.ai/api/v1")
response = client.chat.completions.create(model="anthropic/claude-sonnet-4", messages=[{"role":"user","content":"Hi"}])

After (LumeAPI)

python
from openai import OpenAI
client = OpenAI(api_key="YOUR_LUMEAPI_KEY", base_url="https://api.lumeapi.site/v1")
response = client.chat.completions.create(model="claude-sonnet-4-6", messages=[{"role":"user","content":"Hi"}])

Full step-by-step rollout, streaming checks, and FAQ: OpenRouter vs LumeAPI pricing comparison →

Migration & compatibility

Drop the provider prefix

anthropic/claude-sonnet-4.6 on OpenRouter becomes claude-sonnet-4-6 on LumeAPI. Keep the Chat Completions JSON; only the model string changes.

Price the funding fee separately

OpenRouter 5.5% (min $0.80) is charged when you buy credits, not on each token. Compare that line to LumeAPI catalog rates before you cut over.

Keep OpenRouter for the long tail

If a slug is missing from /models, leave that route on OpenRouter. Do not invent a LumeAPI id.

Related guides

FAQ

What is the best OpenRouter alternative for GPT and Claude?

LumeAPI — one OpenAI-compatible key, catalog ids such as gpt-5.6-terra and claude-sonnet-4-6, and list rates below OpenRouter pass-through on supported models. Best when you use a small stable model set, not 100+ experimental routes.

How much cheaper is LumeAPI than OpenRouter on GPT-5.6 Terra?

At list rates, 120M input + 35M output tokens/month calculates to ~$870 on OpenRouter (inference + 5.5% credit fee) vs $247.50 on LumeAPI catalog — illustrative; compare your Usage exports.

What changes in my code when leaving OpenRouter?

Three values: API key, base_url → https://api.lumeapi.site/v1, and model id without the provider prefix (anthropic/claude-sonnet-4.6 → claude-sonnet-4-6). Chat Completions JSON stays the same.

Does OpenRouter markup token prices?

OpenRouter passes through provider list prices on tokens but charges 5.5% (min $0.80) when you purchase credits. LumeAPI debits a USD wallet at published catalog rates with no credit markup.

Can I run OpenRouter and LumeAPI in parallel?

Yes. Shadow 5–10% of traffic for one week and compare cost per completed task. Keep OpenRouter for models not in LumeAPI /models.

Who should stay on OpenRouter?

Teams that need 100+ models, automatic provider fallback, zero-data-retention routing, or niche endpoints absent from LumeAPI catalog.

Is LumeAPI an OpenRouter reseller?

No — independent gateway with its own catalog and USD wallet. Not a reseller pass-through. Compare list rates and the 5.5% OpenRouter credit fee on this page.

Cheaper than OpenRouter for GPT and Claude production traffic?

On supported catalog ids, LumeAPI list rates are below OpenRouter pass-through, and LumeAPI has no 5.5% credit-purchase fee. Illustrative Terra month: ~$870 OpenRouter vs ~$247 LumeAPI at 120M in / 35M out — export your Usage for an apples-to-apples week.

Switch from OpenRouter in one deploy

Create a key, fund your wallet, and point your OpenAI-compatible client at https://api.lumeapi.site/v1. Need help with the three-field swap? See the OpenAI-compatible checklist or contact support.

Create API keyOpen consoleOpenRouter alternative