Home/AI/@say_gm_
67
Score · promising
Scored 2mo ago · may be stale

@say_gm_

Good Morning

SayGM is an early-stage AI inference gateway offering drop-in access to frontier models (Claude, GPT, Gemini) and open-weight models through a TEE-verified (Intel TDX) infrastructure. The project is live on mainnet with working API, integrated with Bittensor subnet 28, and emphasizes privacy through confidential computing where prompts are sealed inside hardware enclaves. Account is 3 months old with under 1k followers, demonstrating genuine technical differentiation in the AI inference space through verifiable attestation.

AI Analysispromising

Confidence
75%

SayGM is an early-stage AI inference gateway offering drop-in access to frontier models (Claude, GPT, Gemini) and open-weight models through a TEE-verified (Intel TDX) infrastructure.

The project is live on mainnet with working API, integrated with Bittensor subnet 28, and emphasizes privacy through confidential computing where prompts are sealed inside hardware enclaves.

Account is 3 months old with under 1k followers, demonstrating genuine technical differentiation in the AI inference space through verifiable attestation.

Green flags: Working product with live API and real integrations (Cursor, Cline, Vercel AI SDK) · Novel technical approach using Intel TDX TEEs for verifiable privacy, not just policy promises · Early-stage metrics: 848 followers, 3-month-old account, growing user base week-over-week · Connected to Bittensor subnet 28, adding credibility through established infrastructure · First to serve cutting-edge models like Ornith-1.5-397B according to recent activity

Red flags: Single Tier B scout with modest track record (52/100) · Limited social proof beyond Bittensor connection · Competitive space with established players (OpenRouter, LiteLLM) requiring strong differentiation to survive

Token
No · pre-launch
Chain
Bittensor
Stage
mainnet+live
Category
AI inference gateway

Recent tweetsSee all on 𝕏 →

The models you want. For less. With privacy. Running 38% cheaper on average than competitors, and you don't have to sacrifice your privacy. https://t.co/Rh7GDDKgcP
1w ago♥ 67💬 1🔁 12
Using Claude, GPT or Gemini with client details in your prompts? Turn on guardrails in SayGM desktop. Detected emails and phone numbers are swapped for placeholders before the model sees them, and restored in your reply. https://t.co/9uxLaNgIoy
1w ago♥ 55💬 3🔁 8
Lately you’ve mostly heard two things from us: new models and discounts. More models to try, lower prices on the ones you already use. But there’s another reason to route through SayGM: we can’t read your prompts. Our gateway runs inside an Intel TDX enclave, a sealed part of the processor. Your request is only decrypted inside it, and our team and the machine’s host sit outside the seal. Pick a confidential model (the ones ending in -tee) and your prompt stays sealed while the model runs too, so the provider running it can’t read it either. And anyone can check our setup. The hardware produces a signed proof of what the gateway is running, no API key needed. Now think about who needs this. A law firm summarizing a client’s contract. A clinic drafting notes from a patient visit. A bank’s analysts working through financial records. An HR team reviewing a complaint. A startup pointing a coding agent at unreleased code. All of them want what AI can do. None of them can be casual about where that data goes. These are some of the biggest industries in the world, and they all start with the same question: who can read this? We built SayGM so that answer is something you can verify.
2w ago♥ 87💬 3🔁 17
How come you don't use @say_gm_ ?
2w ago♥ 39💬 3🔁 3
Mistral 4 Large live on SayGM. Mistral released the weights yesterday. You can use it on SayGM today. It's Mistral's new open-weight flagship, served by @KubeTEEAI inside a TEE, so your prompts stay sealed end to end. https://t.co/VgJYZn1WgT
2w ago♥ 55💬 3🔁 6

Signal Timeline

0X
@0xKant_ followed
BFirst discovered·2mo ago

Score breakdown0–100

🎯Scout conviction
+18.2 / 35
📚Scout consensus
0 / 10
🪪Profile & earliness
+17 / 20
✍️Substance
+16 / 20
🤖AI verdict
+21.6 / 30
⚠️Penalties
-6 / 40
67
Below threshold (74)
Watching for additional signals.
Followers
848
Account age
5mo
Scouts
1
First seen
2mo ago