//01  OpenRouter alternative

Save by switching
from OpenRouter.

OpenRouter charges a fee every time you top up your balance. SayGM routes to the same frontier models at or below the price the model maker charges, sealed inside a Trusted Execution Environment. Yours stays yours.

Isometric rendering of two routing chips, the lit SayGM chip and a dim one carrying the OpenRouter mark, joined by a track carrying a signal into the lit connector
Typical saving · 9 models28%below the same model on OpenRouter, once their 5.5% credit fee is counted.
AvailableNative OpenAI, Anthropic and Gemini APIs
[02] The price

At or below list price,
by design.

SayGM prices every model at or below the model maker’s published list price, at any volume and from your first request. It is how the supply works: independent operators compete to serve each request, and that competition sets the rate.

Funding your balance costs the amount you fund it with. What the table says is what you are billed.

OpenRouter• List price + 5.5% on credit top-ups (5% via crypto)Inference is passed through at the provider’s own rate.
SayGMAt or below list · top-ups credited in fullCapped at the model maker’s list price.
[03] The prompt

“Where does this data go?” needs a better answer than a privacy policy.

If you route health records, legal documents or source code through an API, a privacy policy is a promise about what a company will choose to do. SayGM’s gateway runs inside an Intel TDX trusted execution environment, and the attestation quote proving it is independently verifiable — so the answer to a security review is something your team checks for itself. The full argument is in hardware attestation vs a privacy policy.

The provider serving the model still sees your prompt in plaintext, exactly as it would if you called it directly. The SayGM gateway in between runs sealed, so your prompt stays out of reach of its operators and the host machine.

Verifiable by anyone

Remote attestation lets you confirm for yourself exactly what code the gateway is running.

Sealed from the routing layer too

Most privacy claims stop at the model. This one covers the gateway itself.

Built for the audit trail

Pairs with signed logs of which model handled what, for teams that need to show their work.
[04] Calculator

What you could save.

Pick a model, set your monthly volume, see the delta.
Model

Estimate using a recently reported price.

50M
15M
91%

Most input tokens in real traffic are cache reads, and both gateways price them well below a fresh one — so the share matters more to the bill than the totals do. The default is what SayGM measures across its own traffic.

SayGM$374.40
OpenRouter$444.31
You keep$69.91per month · 15% lower$838.93 over a yearstart saving
[05] Model by model

Price per million tokens.

USD per 1M tokens, uncached input / output. Reported SayGM rates were reported as of 56 minutes ago.OpenRouter’s is the rate their own public model API publishes, plus the 5.5% credit fee, as of 3 hours ago.
ModelSayGMOpenRouterYou save
Claude Opus 5Anthropic$4.45 / $22.23$5.28 / $26.3815%
Claude Sonnet 5Anthropic$1.78 / $8.89$2.11 / $10.5515%
Claude Haiku 4.5Anthropic$0.89 / $4.45$1.06 / $5.2815%
GPT-5.5OpenAI$3.75 / $22.50$5.28 / $31.6528%
GPT-5.6 TerraOpenAI$1.50 / $9.00$2.11 / $12.6628%
GPT-5.6 LunaOpenAI$0.15 / $0.90$0.21 / $1.2728%
Gemini 3.1 Pro PreviewGoogle$1.40 / $8.40$2.11 / $12.6633%
Gemini 3.5 FlashGoogle$1.05 / $6.30$1.58 / $9.5033%
GLM-5.3Z.ai$0.75 / $2.36$0.04 / $3.5829%
browse the full catalogue
[06] Side by side

SayGM vs. OpenRouter, line by line.

DimensionSayGMvs. OpenRouter
Prompt privacyGateway sealed in an Intel TDX TEE, out of reach of its operators and host machinesPrompts are processed by OpenRouter’s gateway under its published privacy policy
Privacy guaranteeHardware-attested and independently verifiablePolicy commitment, honoured by the company
Confidential open modelsOpen-weights models served inside a TEE end to end, gateway includedOpen models from TEE hosts such as Phala, reached through OpenRouter’s gateway
PricingAt or below the model maker’s list price, top-ups credited in fullProvider pricing passed through, plus a credit-purchase fee
Model coverageClaude, GPT, Gemini and a curated set of open modelsSeveral hundred models — a far broader catalogue
RoutingBuild cascade models that try the next model when one fails, or fusion models that combine several. Every request is balanced across providers by latency, reliability and price, retried on another provider if an attempt fails, and kept on one provider through a conversation where it can, so prompt caching holdsAutomatic retry on the next provider after an error, configurable provider routing, and model variants that sort providers by price or throughput
API surfaceEach model on its own native API, so provider-specific features stay availableOne OpenAI-shaped surface across every model, so a single client covers the whole catalogue
SayGM lists 65 models today. If the one you need is not among them, OpenRouter is the better tool for that job — and running both is a reasonable answer, since a second route to the same models is worth having whichever you make primary.
[07] Questions

Common questions about switching.

Is OpenRouter reliable?

We have not measured it, and a competitor is the wrong source for that number — check their status page and judge for yourself. What we would say is that a second route to the same models is worth having whichever gateway you make primary, and plenty of teams run both and fail over between them.

Is OpenRouter worth it?

For catalogue breadth, yes. It is a credible product carrying several hundred models, far more than SayGM. If your requirement is a gateway whose operator is locked out of your prompts by hardware, that is a different guarantee, and it is the one SayGM is built around. Weighing up more than these two? Our roundup of OpenRouter alternatives compares five of them on price and privacy.

Is OpenRouter free?

They publish a free tier alongside paid usage, and the current terms are in their own docs. Their paid model, as they document it, passes each provider’s per-token rate through with no markup on inference and charges a fee when you buy credits — 5.5% (minimum $0.80) by card, 5% by crypto at the time of writing. SayGM is pay-as-you-go from a prepaid balance, every top-up is credited in full, and every model is priced at or below the model maker’s own list price.

Does OpenRouter have rate limits?

Yes — they vary by tier, the free tier most of all, as set out in their own docs. SayGM’s limits are set per key and lifted on request.

Does OpenRouter save my chats?

Not by default — logging is opt-in, and their published policy on it is reasonable. The difference is where the guarantee lives: SayGM’s gateway runs inside an Intel TDX enclave, so the hardware seals your prompt from the people operating it, on every request, whatever anyone configures.

How much cheaper is SayGM?

Across the models this page compares, 28% for the middle one, counting OpenRouter’s credit-purchase fee. It varies by model, and the calculator above prices your exact volumes. Some models in SayGM’s wider catalogue cost more than they do on OpenRouter — where our own list price sits above theirs, the provider discount does not always close the gap.

Ready to switch?

Not sure OpenRouter is the one to leave? Compare the other OpenRouter alternatives.