Say gm tofrontier models, confidential AI, private routing and wallet payments.

An LLM gateway for Claude, GPT, Gemini and open models.

Up to 70% cheaper, sealed in a gateway you can verify.

Works with Cursor, Cline, Claude Code and the AI SDK

Point Cursor, Cline, Claude Code, or your AI SDK at the SayGM gateway. One base URL, and your existing code keeps working.

View all integrations →

VS Code → Cline settings → Anthropic

API ProviderAnthropicUse custom base URLhttps://api.saygm.com/v1API Keygm_live_…Model IDclaude-opus-5-5
SAYGM DESKTOP

Now on your desktop.

An AI agent that does real work on your files, with the newest models from different labs. Pay only for what you use.

  • Works on your files and folders
  • Switch models mid-chat
  • Client details stay sealed
SayGM Desktop home screen, ready to start a task

Access you can build on.

See live pricing
//01

Models stay reachable

routed across competing operators

//02

Rates in the open

published live, capped at list price

//03

Integrations keep working

native OpenAI, Anthropic and Gemini APIs

Frontier models

Sealed on the way to the model.

Claude, GPT and Gemini run on their provider’s hardware, under its API terms. The SayGM gateway seals your request until it reaches them, and guardrails can swap out personal data on the way and restore it in the reply.

YouSayGM routerModel
Confidential models

Your prompt stays inside the enclave, end to end.

Open-weight models served inside a TEE, so your prompt is decrypted only within the enclave that runs the model. The strongest tier: sealed from request to response.

YouSayGM routerModel
How SayGM's private AI inference works →
Pay as you goNo top-up feesCheaper than list price
Attestation

Verified by hardware. Check it yourself.

The gateway runs in an Intel TDX enclave. The chip itself signs a measurement of exactly what's running, so the proof comes straight from Intel's silicon and anyone can verify it.

What is private AI inference? →
Checking TEE attestation

Reading hardware-signed evidence…

Pricing

Popular models, live pricing.

Per million tokens in USD. Prices use a recent finalized result when available, with eligible offers from miners now shown for models without one.

26%Average reported discount from recent epochs
LiveProvider bidding
63Models
≤ ListCapped at the maker's price
InOut
Claude Opus 5.5claude-opus-5-5Anthropic-11.0%$4.00$3.56$20.00$17.80
GPT 6 Lunagpt-6-lunaOpenAI-23.0%$0.10$0.08$0.50$0.39
Kimi K3kimi-k3OpenMoonshot AI-66.0%$3.00$1.02$15.00$5.10
GLM 5.3 Flashglm-5.3-flashOpenZ.ai-10.3%$0.15$0.13$0.50$0.45
Claude Sonnet 5claude-sonnet-5Anthropic-11.0%$2.00$1.78$10.00$8.90
Qwen3.8 27Bqwen3.8-27bOpenQwen-91.0%$0.50$0.05$3.00$0.27
Claude Fable 5.1claude-fable-5-1Anthropic-11.0%$10.00$8.90$50.00$44.50
GPT 6.1 Solgpt-6.1-solOpenAI-23.0%$2.00$1.54$10.00$7.70
GPT 5.6 Solgpt-5.6-solOpenAI-23.0%$4.00$3.08$20.00$15.40
Claude Sonnet 5.5claude-sonnet-5-5Anthropic-11.0%$2.00$1.78$10.00$8.90
See all 64 models and prices →

Ready to switch?