SuperAPI AI Directory
Partner with us中文AI newsFavorites
All toolsAI developer platformsSayGM

SayGM

SayGM is an LLM gateway service that provides unified access to Claude, GPT, Gemini, and open-source models. It is compatible with the OpenAI or Anthropic SDK, claims prices can be as low as within 30% off list price, and can obtain an API key in about one minute. The gateway supports routing requests to competing operators, with rates publicly disclosed and no higher than list price. On security, the SayGM gateway uses an Intel TDX confidential computing environment, and the hardware signature can be externally verified; for open-weight models, inference is decrypted within a TEE, striving to keep the entire process from request to response sealed. It also offers a desktop AI agent that can handle tasks based on files and folders, supports switching models mid-chat, and charges by usage.

Open website

About this tool

Purpose

SayGM is an LLM (large language model) gateway for developers, used to uniformly call Claude, GPT, Gemini, and open-source models. Its goal is to connect to existing code with one base URL, be compatible with the OpenAI or Anthropic SDK, and claim to provide rates no higher than list price and up to about 70% cheaper.

Main capabilities

  • Multi-model access: Covers Claude, GPT, Gemini, and open-source models, and model requests can be routed to competing operators.
  • SDK compatibility: Can be used with the OpenAI, Anthropic, and Gemini native APIs, and can be used with Cursor, Cline, Claude Code, and AI SDK.
  • Public pricing: USD pricing per million tokens is published in real time and does not exceed list price; models without priced results will show qualified quotes from miners.
  • Security and sealing: The gateway runs in an Intel TDX confidential computing environment, and the chip signs and measures what is actually running for external verifiers to check. Claude, GPT, and Gemini run on their own hardware under their providers' API terms, and the SayGM gateway injects requests until they reach the provider; guardrails can replace personal data along the way and restore it in replies. Open-weight models are served within a TEE, and prompts are decrypted only within the enclave running the model.
  • Desktop AI agent: Can perform actual tasks on user files and folders, uses the latest models from different labs, supports switching models mid-chat, charges by usage, and states that client details remain sealed.

How to use

  1. Get an API key from SayGM (officially stated to be completed in about one minute).
  2. Point the base URL of existing code or tools (Cursor, Cline, Claude Code, AI SDK, etc.) to the SayGM gateway.
  3. For example, in VS Code, go to Cline settings and select Anthropic.
  4. On desktop, you can use its AI agent to handle local files, billed by actual usage. The official site also provides a verification entry point: you can have the AI you already use investigate SayGM, question its claims, and compare it with your existing calling method; the relevant button opens your own account in a new tab with the question pre-filled.

Applicable scenarios

Suitable for developers and teams that need unified accessto multiple models, want to reduce calling costs, or have requirements for request sealing and hardware verifiability; also suitable for users who want to use an agent on the desktop to handle local files and pay by usage.

Limitations and not yet provided

  • Specific price figures for particular models, rate limits, account plans, SLAs, available regions, and the specific list of supported open-source models are not yet provided in the materials.
  • The specific operating steps and verification tool details for hardware verification are not yet provided in the materials.
  • The platforms, file types, and security boundaries supported by the desktop agent are not yet provided in the materials.

Related tools

Browse category