cloverfieldParadies
335 posts
Here comes AutoClaw. We offer a new solution to run OpenClaw locally on your own machine.
- Download and start immediately. No API key required.
- Bring any model you like, or use GLM-5-Turbo, optimized for tool calling and multi-step tasks.
- Fully local. Your data never leaves your machine.
We're giving data control back to Claw users.
Meet AutoClaw → autoglm.z.ai/autoclaw/
Join the conversation → discord.gg/jvrbCRSF3x
English
@mweinbach @FireworksAI_HQ @lumendriada Hmm ok not for me then i can get attached to a model thanks for your input
English
from the way i read it, they may remove it next week but the week you paid for you still get. so if you start monday and on tuesday they remove it, you still have it for the rest of that week
i assume they will say "hey in 1-2 weeks we're removing XYZ" but it'll be based on when you sign up, so the removal date would be different?
your guess is as good as mine, seems like they're still figuring it out and offering kimi for now. it's a week by week basis, so whatever you have when you refresh, you have through the week. it's not that you'd lose access you could just not renew for that week
English
@lumendriada I requested access to the thing in the panel then it was available today when I checked
Idk how they handle it
English
@TheGreatAxios @FireworksAI_HQ @Kimi_Moonshot @SkaleNetwork So once a model is in firepass list they wont remove it next week right? Asking since if they add glm 5.1 ill want to keep using it and not lose access in a week
English
Been having a blast playing around with @FireworksAI_HQ Coding Pass and their @Kimi_Moonshot Kimi K2.5-Turbo model.
This was my primary driver late last week and was incredibly capable with helping me scaffold out MPP on @SkaleNetwork + a BITE-focused experimental wallet.
Next Up: BITE + x402, then BITE + ERC-8183
English
@mweinbach @FireworksAI_HQ @lumendriada So once a model is in firepass list they wont remove it next week right? Asking since if they add glm 5.1 ill want to keep using it and not lose access in a week
English
@FireworksAI_HQ @mweinbach @lumendriada Will u suddenly remove the existing model like kimi k2.5 turbo?
English
@mweinbach @lumendriada Open sign up for now, but we have limited capacity. We have a little progress bar on the site that shows you availability.
English
@mweinbach Will they suddenly remove the kimi turbo model?
English
Fireworks AI fire pass is so good
It's Kimi K2.5 Turbo right now at like 250 tok/s and idk what the limits are but it's HIGH
oh and free trial but $7 per week
English
@mweinbach But its not working
{
"error": {
"code": "invalid_request_error",
"type": "invalid_request_error",
"object": "error",
"message": "JSON Schema not supported: could not translate the enum None."
},
"endpoint": "https://api...
}
English
My first test of MiniMax M2.7
this made 3d fibre physics with all the specs of m2.7 is written
MiniMax (official)@MiniMax_AI
Early testers are saying that M2.7 has big improvements in emotional intelligence and character consistency 👀
English
@F2aldi The model is probably quantized hard and will feel stupider
English
Wow, this is crazy.
For $10, you get access to GLM-5, Kimi K2.5, and MiniMax M2.5.
And right now they increased the limit by 3×, which is really nice.
Every 5 hours, you get roughly $12 worth of usage, which is honestly pretty good.
dax@thdxr
we've increased opencode go's limits by 3x - still $10/month
English
@TukiFromKL If 16m request is enough to distilled their model then the chinese has reach breakthrough in distillation techniques lol
English
Anthropic scraped the entire internet to build Claude.
Now they're mad someone scraped Claude to build their own. The irony is so thick you could train a model on it.
Anthropic@AnthropicAI
We’ve identified industrial-scale distillation attacks on our models by DeepSeek, Moonshot AI, and MiniMax. These labs created over 24,000 fraudulent accounts and generated over 16 million exchanges with Claude, extracting its capabilities to train and improve their own models.
English
@ai_hyperbull @ArthurB If 16m request is enough to distilled their model then the chinese has reach breakthrough in distillation techniques lol
English
Horrible arguement. They're not giving me a copy of your book through Claude, they're using your book to predict my next token. Huge difference.
As for the Chinese, they're taking Claude, cloning him, and reselling for 10x cheaper because they didn't have to pay for the training... that's unfair. And if you ever ever hope to get paid for the original source material, you better hope it's Anthropic and not China
Anyone using Chinese AI in the US is part of the problem. POTUS should ban Chinese AI form any access to US Infrastructures. They cannot operate here as stolen property. Shut down all Chinese API access in the US
English
Taking terabytes of data without compensating copyright owners is fair use, but using an publicly available AI model to train another one is a distillation attack?
Anthropic@AnthropicAI
We’ve identified industrial-scale distillation attacks on our models by DeepSeek, Moonshot AI, and MiniMax. These labs created over 24,000 fraudulent accounts and generated over 16 million exchanges with Claude, extracting its capabilities to train and improve their own models.
English
@ArthurB If 16m request is enough to distilled their model then the chinese has reach breakthrough in distillation techniques lol
English
@Dr_Gingerballs If 16m request is enough to distilled their model then the chinese has reach breakthrough in distillation techniques lol
English
It is beyond hilarious that competitors exploited their discount pricing model to extract their product at a steep discount and directly compete. And they can’t sue them because their own theft of humanity’s IP will come out in discovery. True chef’s kiss stuff here.
Anthropic@AnthropicAI
We’ve identified industrial-scale distillation attacks on our models by DeepSeek, Moonshot AI, and MiniMax. These labs created over 24,000 fraudulent accounts and generated over 16 million exchanges with Claude, extracting its capabilities to train and improve their own models.
English
@TommyFalkowski If 16m request is enough to distilled their model then the chinese has reach breakthrough in distillation techniques lol
English
Words cannot express how fucking angry this kind of statement is making me. This kind of of psychopathic public behavior should not be condoned for even one second. Are you fucking kidding me? Who even signed off on publishing something like this??? The only chance anthropic and the other closed labs have of redeeming themselves is open sourcing everything they have ever done. Fucking Hypocrites...
Anthropic@AnthropicAI
We’ve identified industrial-scale distillation attacks on our models by DeepSeek, Moonshot AI, and MiniMax. These labs created over 24,000 fraudulent accounts and generated over 16 million exchanges with Claude, extracting its capabilities to train and improve their own models.
English
@TheAhmadOsman If 16m request is enough to distilled their model then the chinese has reach breakthrough in distillation techniques lol
English
a reminder that Anthropic is a
> fear-mongering company thatʼs
> lobbying against opensource AI
> to stop you from running
> your own AI models
theyʼre
> pro-regulation with an agenda
> pushing “safety” as control
> wants to gatekeep, not protect
> malicious
DO NOT TRUST THEM
English
@bdsqlsz If 16m request is enough to distilled their model then the chinese has reach breakthrough in distillation techniques lol
English
So anyone who uses the internet can accuse you of collecting their data for pre-training, a pre-training attack?🤡
Anthropic@AnthropicAI
We’ve identified industrial-scale distillation attacks on our models by DeepSeek, Moonshot AI, and MiniMax. These labs created over 24,000 fraudulent accounts and generated over 16 million exchanges with Claude, extracting its capabilities to train and improve their own models.
English
@synthwavedd If 16m request is enough to distilled their model then the chinese has reach breakthrough in distillation techniques lol
English
@Teknium If 16m request is enough to distilled their model then the chinese has reach breakthrough in distillation techniques lol
English
Anthropic’s endless need to be a whiny asshole has finally culminated in all of twitter becoming unified for a day
English
@Teknium I don't get it. Obviously they would be frustrated that everyone is distilling their models. Distilling opus is an easy option for everyone except the person building opus
English
Complete list of models currently worth using for code:
Opus 4.6
Codex 5.3
GLM-5
English