7 ms·

It could be but there are a host of companies going after open weights models: Arcee, Reflection, Llama (TBD on Meta's focus on closed-source versus open-source

by gkapur 2mo ago

It could be but there are a host of companies going after open weights models: Arcee, Reflection, Llama (TBD on Meta's focus on closed-source versus open-source), etc. That said, the fine-tuning API + open weight model at least is a semblance of a viable business that could work so I will be curious about it. I'm not sure the synergy is fully there (why is someone with an open weights model privelaged to fine-tune it better if it's just QLora or Lora) but let's see!

paxys 2mo ago
Llama is dead. Meta is now releasing proprietary models (Muse Spark).
- anon373839 2mo ago
  They’ve made some wishy-washy statements about their intention to release a future version of Muse Spark as open weights. We’ll see.
andriy_koval 2mo ago
> It could be but there are a host of companies going after open weights models: Arcee, Reflection, Llama (TBD on Meta's focus on closed-source versus open-source), etc. my bet is that Chinese government fund Chinese models way more compared to what those companies receive (except llama, which is outdated but was strong foundation at its time)
- mannanj 2mo ago
  I have a similar bet. Looks like people don't like this idea. You got downvoted a lot.
  - sgt101 2mo ago
    Some stories recently that China has decided to reign in and manage it's model production landscape much more tightly.
    - mannanj 2mo ago
      Ah. Is it true? What do you think?
      - sgt101 2mo ago
        I think... we will see!
- gkapur 2mo ago
  The story of Reflection AI is supposedly that the company was faffing and failing at winning in the coding agent space, but was introduced to Jenson, who suggested they build an open-weight model and said he would fund it. That turned into a $2 billion financing with NVIDIA doing roughly $500 million and was a complete pivot. I think the bet would have to be that a US Open Weight company either: 1. Gets a lot of money from Jenson who views them as a counterbalance to the big labs in his ecosystem and a way to generate leverage (the same way he is positioning neoclouds-- it also could be synergistic with neoclouds who could offer the model serving endpoints) 2. Can fast follow the same way Mistral does (which, honestly, seems like just distilling the Chinese model, which distills the US lab but is pretty innovative on a whole lot of architecture both in training and serving land.) 3. AND figure out some (maybe not super lucrative but lucrative enough) sort of business model, as well. There are lots of possible business models, so I will be curious how this whole space evolves.
  - htrp 2mo ago
    >The story of Reflection AI is supposedly that the company was faffing and failing at winning in the coding agent space, but was introduced to Jenson, who suggested they build an open-weight model and said he would fund it. That turned into a $2 billion financing with NVIDIA doing roughly $500 million and was a complete pivot. You can pretty much remove the supposedly here
  - andriy_koval 2mo ago
    > That turned into a $2 billion financing with NVIDIA doing roughly $500 million and was a complete pivot. I suspect 2B is not enough to boostrap frontier model from the scratch (for both talent and hardware)
  - fmajid 2mo ago
    Jensen Huang is just trying to commoditize the complements to his GPUs.
    - drob518 2mo ago
      Cf Microsoft v Intel circa 1995
YetAnotherNick 2mo ago
Do any of these even have match a year old Deepseek 3.1?
- suprjami 2mo ago
  DS3 isn't even looked at anymore. GLM-5.2 is the best in that class right now. It is competitive with current GPT/Claude/Gemini.
  - blovescoffee 2mo ago
    "Current GPT/Claude/Gemini" is not a meaningful statement about perf. There's many different models from each of those providers and there's a considerable gap between the best of anthropic and open ai compared to gemini. Benchmarks have GLM 5.2 somewhere underneath Sol and Fable and closer to now last-gen openai and anthropic models.
    - spwa4 2mo ago
      One error: GLM 5.2 beats the best public Gemini model, 3.5 pro. There's 2 caveats with the rest. First, GLM 5.2 matches those models in "xhigh" effort modes, which has a very low quota on the subscriptions, especially for Claude. Second, last-gen GPT/Claude means what they release in April/May of 2026. Or to be even more complete/fair: GLM 5.2 beats what OpenAI released in March 2026 (GPT 5.5 xxhigh), and what Anthropic released in April 2026 (Opus 4.7 xhigh). It is beaten by what OpenAI released in April of 2026 (GPT 5.6 Sol xxhigh) and Anthropic released in May 2026 (Opus 4.8 (the same as "Fable" ?), xhigh effort) GLM 5.2 was released on Jun 16 and if OpenAI and Anthropic hadn't done those quick releases they would have been beaten on their best available models ... So great news! Open source now has SOTA performance 3 months after OpenAI/Anthropic/Google. Wow.
      - suprjami 2mo ago
        The gap has been steadily closing over time. Opus 4.8 (May) to Kimi K3 (July) has apparently just dropped it to two months. China also does efficiency improvements. Qwen 3.6 27B is better than Sonnet 4.5 and you can run it on a couple of gaming video cards. That's incredible. I can do real actual work with this! As Google said in 2023, none of them have a moat, open weight models will win.
jjfoooo4 2mo ago
I don’t really get the business plan for open weights model companies, is the idea companies would pay them for serving?
- tfehring 2mo ago
  Thinky's main commercial product AFAIK is Tinker [0] - companies pay them to host their fine-tuning workloads and then the resulting fine-tuned models. I don't know if this is a good business plan, but I'm sure at least one person there has read Joel on Software [1]. [0] https://thinkingmachines.ai/tinker/ https://thinkingmachines.ai/tinker/ [1] https://www.joelonsoftware.com/2002/06/12/strategy-letter-v/ https://www.joelonsoftware.com/2002/06/12/strategy-letter-v/
  - mchusma 2mo ago
    I don’t know if it’s a great business model but it makes perfect sense to me. Open models when fine tuned are capable at better than frontier performance at a fraction of the price for many (probably most) domain specific tasks. If companies help make that easy to implement, there is value to capture. But I kind of like Unsloths model here which is to be really good at just layer, and not bothering with building their own models.
  - sgt101 2mo ago
    I don't get this - I can do LORA on my mac... ok I can't do LORA on a 1tn param model, but if I was in the tn parameter model game I would get some kit that I could use to do that... What's their moat / secret sauce?
    - tfehring 2mo ago
      Like, buy and set up the physical hardware? I cba with that. Plus the hardware you want for LoRA (the type but especially the quantity) is different than what you want for inference, so either you'd under-spec it and wait forever for fine-tuning runs, or over-spec it and have low utilization most of the time. And even then who knows if it would be good enough to LoRA next year's best open source model. AWS gets great margins for renting out commodity hardware as a service because it built the right abstractions and can serve them efficiently at scale, I think the arguments here are basically the same.

by gkapur 2mo ago

paxys 2mo ago

Llama is dead. Meta is now releasing proprietary models (Muse Spark).

anon373839 2mo ago
They’ve made some wishy-washy statements about their intention to release a future version of Muse Spark as open weights. We’ll see.

andriy_koval 2mo ago

> It could be but there are a host of companies going after open weights models: Arcee, Reflection, Llama (TBD on Meta's focus on closed-source versus open-source), etc. my bet is that Chinese government fund Chinese models way more compared to what those companies receive (except llama, which is outdated but was strong foundation at its time)

mannanj 2mo ago
I have a similar bet. Looks like people don't like this idea. You got downvoted a lot.
- sgt101 2mo ago
  Some stories recently that China has decided to reign in and manage it's model production landscape much more tightly.
  - mannanj 2mo ago
    Ah. Is it true? What do you think?
    - sgt101 2mo ago
      I think... we will see!
gkapur 2mo ago
The story of Reflection AI is supposedly that the company was faffing and failing at winning in the coding agent space, but was introduced to Jenson, who suggested they build an open-weight model and said he would fund it. That turned into a $2 billion financing with NVIDIA doing roughly $500 million and was a complete pivot. I think the bet would have to be that a US Open Weight company either: 1. Gets a lot of money from Jenson who views them as a counterbalance to the big labs in his ecosystem and a way to generate leverage (the same way he is positioning neoclouds-- it also could be synergistic with neoclouds who could offer the model serving endpoints) 2. Can fast follow the same way Mistral does (which, honestly, seems like just distilling the Chinese model, which distills the US lab but is pretty innovative on a whole lot of architecture both in training and serving land.) 3. AND figure out some (maybe not super lucrative but lucrative enough) sort of business model, as well. There are lots of possible business models, so I will be curious how this whole space evolves.
- htrp 2mo ago
  >The story of Reflection AI is supposedly that the company was faffing and failing at winning in the coding agent space, but was introduced to Jenson, who suggested they build an open-weight model and said he would fund it. That turned into a $2 billion financing with NVIDIA doing roughly $500 million and was a complete pivot. You can pretty much remove the supposedly here
- andriy_koval 2mo ago
  > That turned into a $2 billion financing with NVIDIA doing roughly $500 million and was a complete pivot. I suspect 2B is not enough to boostrap frontier model from the scratch (for both talent and hardware)
- fmajid 2mo ago
  Jensen Huang is just trying to commoditize the complements to his GPUs.
  - drob518 2mo ago
    Cf Microsoft v Intel circa 1995

YetAnotherNick 2mo ago

Do any of these even have match a year old Deepseek 3.1?

suprjami 2mo ago
DS3 isn't even looked at anymore. GLM-5.2 is the best in that class right now. It is competitive with current GPT/Claude/Gemini.
- blovescoffee 2mo ago
  "Current GPT/Claude/Gemini" is not a meaningful statement about perf. There's many different models from each of those providers and there's a considerable gap between the best of anthropic and open ai compared to gemini. Benchmarks have GLM 5.2 somewhere underneath Sol and Fable and closer to now last-gen openai and anthropic models.
  - spwa4 2mo ago
    One error: GLM 5.2 beats the best public Gemini model, 3.5 pro. There's 2 caveats with the rest. First, GLM 5.2 matches those models in "xhigh" effort modes, which has a very low quota on the subscriptions, especially for Claude. Second, last-gen GPT/Claude means what they release in April/May of 2026. Or to be even more complete/fair: GLM 5.2 beats what OpenAI released in March 2026 (GPT 5.5 xxhigh), and what Anthropic released in April 2026 (Opus 4.7 xhigh). It is beaten by what OpenAI released in April of 2026 (GPT 5.6 Sol xxhigh) and Anthropic released in May 2026 (Opus 4.8 (the same as "Fable" ?), xhigh effort) GLM 5.2 was released on Jun 16 and if OpenAI and Anthropic hadn't done those quick releases they would have been beaten on their best available models ... So great news! Open source now has SOTA performance 3 months after OpenAI/Anthropic/Google. Wow.
    - suprjami 2mo ago
      The gap has been steadily closing over time. Opus 4.8 (May) to Kimi K3 (July) has apparently just dropped it to two months. China also does efficiency improvements. Qwen 3.6 27B is better than Sonnet 4.5 and you can run it on a couple of gaming video cards. That's incredible. I can do real actual work with this! As Google said in 2023, none of them have a moat, open weight models will win.

jjfoooo4 2mo ago

I don’t really get the business plan for open weights model companies, is the idea companies would pay them for serving?

tfehring 2mo ago
Thinky's main commercial product AFAIK is Tinker [0] - companies pay them to host their fine-tuning workloads and then the resulting fine-tuned models. I don't know if this is a good business plan, but I'm sure at least one person there has read Joel on Software [1]. [0] https://thinkingmachines.ai/tinker/ https://thinkingmachines.ai/tinker/ [1] https://www.joelonsoftware.com/2002/06/12/strategy-letter-v/ https://www.joelonsoftware.com/2002/06/12/strategy-letter-v/
- mchusma 2mo ago
  I don’t know if it’s a great business model but it makes perfect sense to me. Open models when fine tuned are capable at better than frontier performance at a fraction of the price for many (probably most) domain specific tasks. If companies help make that easy to implement, there is value to capture. But I kind of like Unsloths model here which is to be really good at just layer, and not bothering with building their own models.
- sgt101 2mo ago
  I don't get this - I can do LORA on my mac... ok I can't do LORA on a 1tn param model, but if I was in the tn parameter model game I would get some kit that I could use to do that... What's their moat / secret sauce?
  - tfehring 2mo ago
    Like, buy and set up the physical hardware? I cba with that. Plus the hardware you want for LoRA (the type but especially the quantity) is different than what you want for inference, so either you'd under-spec it and wait forever for fine-tuning runs, or over-spec it and have low utilization most of the time. And even then who knows if it would be good enough to LoRA next year's best open source model. AWS gets great margins for renting out commodity hardware as a service because it built the right abstractions and can serve them efficiently at scale, I think the arguments here are basically the same.