Lewbar

[Lewbar @Lewbare]

@steipete @JosephJacks_ @davemorin Use Chutes - integrates directly into Openclaw and no one can see your data

[Chutes @chutes_ai]

You can now run frontier AI models where not even the gpu provider can see your data.

15+ models with hardware-enforced privacy (TEE) on Chutes.

No other open-source inference provider offers this.

Here's the full lineup and why it matters ↓

[Lewbar @Lewbare]

@petergyang Use Chutes - Minimax, GLM and Kimi. No one can see your data, integrates directly into Openclaw

[Chutes @chutes_ai]

You can now run frontier AI models where not even the gpu provider can see your data.

15+ models with hardware-enforced privacy (TEE) on Chutes.

No other open-source inference provider offers this.

Here's the full lineup and why it matters ↓

[Lewbar @Lewbare]

@cathrynlavery @openclaw Use Chutes - Minimax, Kimi, GLM and no one can see your data. Integrates directly into Openclaw

[Chutes @chutes_ai]

You can now run frontier AI models where not even the gpu provider can see your data.

15+ models with hardware-enforced privacy (TEE) on Chutes.

No other open-source inference provider offers this.

Here's the full lineup and why it matters ↓

[Lewbar @Lewbare]

@hyperbeat code please

[Lewbar @Lewbare]

@badenglishtea @jon_durbin This would be amazing….

[Chutes @chutes_ai]

@Kimi_Moonshot Kimi K2.5 just topped the open source leaderboard.
1T parameters. Native multimodal. Agent Swarm that coordinates 100 sub-agents in parallel.
And we're running it inside a Trusted Execution Environment.
Here's why that matters 🧵

[Jon Durbin @jon_durbin]

🔒The full client-side E2E encryption framework is ready to be deployed, along with a boatload of other updates. Hope to start rolling it out tomorrow morning.
https://t.co/OMTC2rExXJ

This will likely be the most secure and verifiable inference anywhere on the planet, with zero (well, let's say infinitesimally small) risk of evesdropping or prompt leakage/etc.

So, TEE node spins up, creates ephemeral quantum-safe encryption key, client gets the quote for instances to verify the secure enclave, gets the public key for an instance, encrypts their request for exactly that one instance and sends along a response encryption key. Only the client and that TEE pod will ever see the request, no evesdropping even possible, and being TEE nodes with in-memory keys no chance of decrypting or seeing the traffic regardless even with physical access to host.

The other PRs:
https://t.co/ba3jtiqjqg
https://t.co/nA6dFT5ybI
(and like 100k lines of C code part of our proprietary aegis library)

The amount of stuff in aegis (and then our proprietary virtualization/obfuscation/packing/crypto/etc. lib) is quite extensive.

Everything is stable in dev 🚀

[Lewbar @Lewbare]

@jaylavee_ @chutes_ai @basilic_ai Has been stable for me now there are model instances running. Did you use the Chutes install script?

[Lewbar @Lewbare]

@thdxr Use Chutes - TEE as well

[Chutes @chutes_ai]

@Kimi_Moonshot's Kimi K2.5 is LIVE on Chutes 🔥

Visual agentic intelligence from Moonshot.
Beating Claude Opus 4.5 on agentic benchmarks.
→ 50.2% HLE (global SOTA)
→ 74.9% BrowseComp (global SOTA)
→ 76.8% SWE-Bench Verified
→ Agent Swarm: 100 parallel sub-agents

Try it now → https://t.co/srzZ7tgrcg

[Chutes @chutes_ai]

We just made OpenClaw stupid easy to set up.
One command. 60+ open source models. Under 60 seconds.
Meet ParaClaw 🧵

[Lewbar @Lewbare]

@Legendaryy @Kimi_Moonshot Use it on Chutes - TEE so your data is secure

[Lewbar @Lewbare]

Using @chutes_ai with @openclaw

✅ Kimi 2.5
✅ TEE security (prompt injection protection)
✅ $3-20/mo flat rate
✅ 10-second OpenClaw setup

Chutes powers @openrouter: 40% of OR traffic

Install:
```bash
curl -fsSL https://t.co/iprJ02U4fD | sh```

#Chutes #OpenClaw #Kimi

[Lewbar @Lewbare]

@Jacobsklug @openclaw @Lovable OS

[Lewbar @Lewbare]

@blocmates Use Chutes - starts at $3/month, 60 models and has TEE

[Chutes @chutes_ai]

ParaClaw spins up @OpenClaw with Chutes in one command.

Give your bot 60+ models in under 60 seconds.

curl -fsSL https://t.co/QOMnfHuyHj | sh

[Lewbar @Lewbare]

@0xzak Use Chutes - starts at $3 / month for 60 models
https://t.co/Dlai8LsmMl

[Chutes @chutes_ai]

ParaClaw spins up @OpenClaw with Chutes in one command.

Give your bot 60+ models in under 60 seconds.

curl -fsSL https://t.co/QOMnfHuyHj | sh

[Lewbar @Lewbare]

@itsgouthamX @heyDhavall Use Chutes with Kimi 2.5 TEE

[Lewbar @Lewbare]

Using @chutes_ai with @openclaw

✅ Kimi 2.5
✅ TEE security (prompt injection protection)
✅ $3-20/mo flat rate
✅ 10-second OpenClaw setup

Chutes powers @openrouter: 40% of OR traffic

Install:
```bash
curl -fsSL https://t.co/iprJ02U4fD | sh```

#Chutes #OpenClaw #Kimi

[Lewbar @Lewbare]

@heyDhavall Chutes offering Kimi 2.5 with TEE at $3/ month

[Lewbar @Lewbare]

Using @chutes_ai with @openclaw

✅ Kimi 2.5
✅ TEE security (prompt injection protection)
✅ $3-20/mo flat rate
✅ 10-second OpenClaw setup

Chutes powers @openrouter: 40% of OR traffic

Install:
```bash
curl -fsSL https://t.co/iprJ02U4fD | sh```

#Chutes #OpenClaw #Kimi

[Lewbar @Lewbare]

@mhdfaran Even better go direct with Chutes (already does a lot of OpenRouter traffic)

[Lewbar @Lewbare]

Using @chutes_ai with @openclaw

✅ Kimi 2.5
✅ TEE security (prompt injection protection)
✅ $3-20/mo flat rate
✅ 10-second OpenClaw setup

Chutes powers @openrouter: 40% of OR traffic

Install:
```bash
curl -fsSL https://t.co/iprJ02U4fD | sh```

#Chutes #OpenClaw #Kimi

[Lewbar @Lewbare]

Using @chutes_ai with @openclaw

✅ Kimi 2.5
✅ TEE security (prompt injection protection)
✅ $3-20/mo flat rate
✅ 10-second OpenClaw setup

Chutes powers @openrouter: 40% of OR traffic

Install:
```bash
curl -fsSL https://t.co/iprJ02U4fD | sh```

#Chutes #OpenClaw #Kimi

[Lewbar @Lewbare]

@felipezb13 @petergyang @openclaw Use Chutes - it’s 3 /month and can set different models do different purposes

curl -fsSL https://t.co/ptH2Mx92vM | sh

[Lewbar @Lewbare]

@Pranit @Legendaryy Use Chutes, TEE as well

[Chutes @chutes_ai]

@Kimi_Moonshot's Kimi K2.5 is LIVE on Chutes 🔥

Visual agentic intelligence from Moonshot.
Beating Claude Opus 4.5 on agentic benchmarks.
→ 50.2% HLE (global SOTA)
→ 74.9% BrowseComp (global SOTA)
→ 76.8% SWE-Bench Verified
→ Agent Swarm: 100 parallel sub-agents

Try it now → https://t.co/srzZ7tgrcg