Private and encrypted. Maple builds your product, audits your code, and scans for vulnerabilities. Privacy by design.
Private AI →
Joined February 2025
- Tweets882
- Following280
- Followers4.3K
- Likes1.3K
Pinned Tweet
New on Maple Max and Team: DeepSeek V4.1 Flash, abliterated.
Refusals removed from the model itself, so it works through security questions instead of declining them.
- 1M context
- Runs in the same attested enclaves as every Maple model
- Built for security research, code audits and red-teaming
Attackers already run uncensored models. Defenders shouldn't be stuck with "I can't help with that."
Big move from Google. Chips built so even an insider can't access your data while AI runs.
This should be the standard. Maple already works this way, and you can verify it yourself.
Today, we're expanding our eighth-generation TPU portfolio with two new systems built specifically for the agentic era — bringing our total to four distinct eighth-generation TPU systems launched in a single year.
TPU 8i Max was built for massive context windows and the highest compute density per rack, and TPU 8tx is optimized for secure and confidential inference. While flexible across workloads, the four systems in our 8th gen TPU portfolio give developers and organizations a purpose-built accelerator for every stage of the agentic lifecycle.
@AminVahdat, SVP and Chief Technologist, AI and Infrastructure, explains why we're expanding our TPU portfolio and what everyone needs to know about TPU 8tx and 8i Max ⬇️
We baked Felix a cake.
🔒🚨 NO NORD NOVEMBER IS HERE 🚨🔒
No Nord November is Obscura’s answer to Black Friday VPN “discounts” that lock you in, only to raise prices later.
NEW: Obscura is now offering seven-day free trials.
Start a 7-day free trial, buy one month, send proof of time left on your current plan, and we’ll match it for free (up to 2 years).
👇 Link
🔗 NoNordNovember.com
Get private superintelligence in your pocket.
10% off your first 3 months of Maple with code FREEFELIX
maple.ai/pricing?utm_source=…
2
6
290
Thanks @TinfoilAI for getting Kimi K3.1 Pro running in secure enclaves so quickly after the weights dropped 🩶 It's live in Maple now.
An open-weight model within a point of Claude Opus 5.5 on agentic coding is now in Maple.
Kimi K3.1 Pro from @Kimi_Moonshot, running in attested secure enclaves, so no third party can read your chats.
In Maple Chat and Maple Work, on Pro, Max and Team.
Near-frontier. Open weights. Private.
What changed from K3, what it takes to run it yourself, and how to use it in Maple 👇
maple.ai/blog/kimi-k3-1-pro-…
2
3
348
Cost is why many teams are moving to open models.
The next question is where the prompts go, and who can read them along the way. That's what attested enclaves are for.
BREAKING: Revolut moved ~2,000 engineers onto self-hosted GLM for coding and pushed Claude Code and Codex to the exception list.
Per David Abitbol, who leads AI ops there: 56 Nvidia B200s, a blended 30 cents per million tokens, roughly a tenth of what they paid Anthropic and OpenAI. Engineering lead time improved 70%.
Here's the part that matters: the gains showed up when tokens stopped being rationed. Engineers quit saving AI for the hard tickets and pointed it at everything.
Frontier models become the exception, open weights become the default. That's a pricing problem for every lab heading to IPO.
h/t @OliverMolander
Would you move your team off frontier APIs if the bill dropped 90%?
Abliterated DeepSeek V4.1 Flash on Maple. Far fewer refusals on security work, privately.
The code you most need a second opinion on is the code you shouldn't paste anywhere. On Maple, it's encrypted before it leaves your device and only decrypted inside secure enclaves, all the way through to the model.
Every entry links to its primary source, and each change shows what the policy said before. It's free to reuse under CC BY 4.0, with RSS, JSON and CSV.
maple.ai/privacy-timeline
198
This is why we build Maple to be private by design, in a way you can check: maple.ai/proof
1
3
239
Local models play a role in the future of sovereign computing. But the "absolute unit" models will likely always need the cloud.
The real question is: how do you run them privately? Secure enclaves you can verify. That's where we come in.
Own your intelligence.
Anyone that opposes this has something to sell you (perhaps your own intelligence back to you.)
Local models need to continue proliferating. The “absolute unit” massive models may be out of reach for now locally, so your best option is to, indeed, keep your options open. Own your harness, data, connections, runtime: your agents. Embrace open source.
1
22
970
When your AI provider can read your sessions, it may end up deciding what you're allowed to work on.
Shame on @DarioAmodei / Anthropic. I never distilled Claude—rollouts came from a strong open-source model; Claude just helped submit jobs and run analysis. Still got tagged for "training a competitor," and now anything touching RFT, SFT, or training larger models gets me blocked, Opus 5.5 downgraded to Opus 5, or the session killed outright. You built your models on the world's data, then treat paying users doing research automation as the enemy. Pathetic.
no application, no asterisk. Security models on Pro, Max and Team, and we don't (can't) train on your chats. maple.ai/cybersecurity
1
231
Avoid this mess and secure your stack with our encrypted DeepSeek abliterated and GLM 5.3.
Maple doesn’t see your prompts or traces, so there’s nothing to train on and nothing to sell to third parties.
Our cyber ready models are available in Maple Chat, Maple Proxy (for connecting to things like OpenCode), and Maple Work (encrypted harness launching next month).
haha, so apparently OpenAI are about to start training on your Daybreak Blue sessions, with no way to opt out.
Literally stealing people's private work, as they do it, is cartoonishly evil.
It also will presumably kill Daybreak usage by security researchers. Or at least, it should. The thing OAI have had to swear they are not doing to math researchers, they are now explicitly about to start doing to security researchers.
Yet another reminder of the importance of open models.