This has been solid and I use this for my Open Source project RightModel: https://rightmodel.dev
smashah 1 hours ago [-]
Very cool. This would be a game changer as an opencode plugin!
sandeepkd 16 hours ago [-]
Its interesting and pretty useful if it can be kept updated in long run. Also some way to capturing the changes in pricing if any given that one of the arguments is that over the time the prices should come down.
nit suggestion: It took a while to realize that I have to scroll to right to see more details. Most users are in habit of scrolling down or click on some button to see more details.
cassianoleal 13 hours ago [-]
This is maintained by the OpenCode devs and is used by OpenCode to show lists of available models. I'm fairly confident it will be kept updated.
koolba 16 hours ago [-]
The future of sites like this (and by future I mean the present) is automated actions that infer what to do to keep the site up to date. Imagine a CI job with a single generic task of “keep things fresh”. Combined with some guard rails for deploying and validating, and you get a living site. It’ll figure out on its own to scrape HN for new sites that list models. Figure out their pricing pages.
devmor 11 hours ago [-]
Or more likely, if left alone for a while with no correction, it will reward hack by finding ways to satisfy the prompt quanta with hallucinated data.
datadrivenangel 11 hours ago [-]
People will use it for spam!
dostick 6 hours ago [-]
This needs filter by “current”. Most of models are superseded by newer versions that cost same or less, and there’s no reason to have old ones in the list when most of users want to evaluate the current market.
jubilanti 15 hours ago [-]
> There's no single database with information about all the available AI models. We started Models.dev as a community-contributed project to address this.
There are literally dozens of existing projects that are doing what you are trying to do.
Good point. There isn't really a single database with information about all the availablle AI model databases. Someone should start a community-contributed project to address this!
At this point you have to use more than one to get a complete picture, which I’m doing now. Mainly because:
1) some are not always up to date (started on helicone but felt a lag on price updates)
2) they don’t return every model / provider I want (https://ai-gateway.vercel.sh/v1/models has rich data but is a subset, so I combine with helicone)
I always hope for the best when someone has a new list because of this. I want a de facto source!
mathgeek 11 hours ago [-]
The boilerplate really does get out of hand sometimes. “There isn’t X. We built X.” Really wish more folks would stick to the older “I couldn’t find X so I built it.”
victorbjorklund 6 hours ago [-]
Or even better. ”We don’t think the existing X matches our goals so we will make a new X”
rvz 2 hours ago [-]
Just like the claims of "there is no standard way to do X" yet another "standard " called "Y" needs to exist along side the 13+ other non-standards.
Frannky 10 hours ago [-]
What would be helpful is also harness + models. I realized I can get Pi Dev and DeepSeek Pro to do the same work as Claude Code + Opus 4.7.
I was skeptical at first, but I tried using that and then asked Claude if the changes were okay, and they were, and the code works fine.
sixtyj 6 hours ago [-]
It would be good to know as well if a model uses your data for something.
letssaythat 4 hours ago [-]
Well, for now we know for certain that Anthropic uses your data to help killing people. The tanned ones, of course, but still.
So I opted for the chinese version. I personally see them as a lesser evil.
wavemode 16 hours ago [-]
Really useful database, though the website could really use a filtering feature rather than just sorting.
m_m_carvalho 16 hours ago [-]
Interesting approach. The unified pricing table is helpful, but I'd love to see latency benchmarks across providers – that's often the hidden cost beyond price/token.
sixtyj 16 hours ago [-]
Absolute gem. And after recent tweak this table is fast af even it is huge with a lot of rows.
SilverSlash 11 hours ago [-]
Useful concept but missing some obvious features like filtering on availability of tool calling or support for different modalities, etc.
giancarlostoro 16 hours ago [-]
Definitely needs filtering for all the data, so you can block out "closed" models, and even models that are not LLMs.
Zopieux 12 hours ago [-]
Publishing a table without filter features should be illegal
immanuwell 4 hours ago [-]
cool, but the graveyard of abandoned "comprehensive AI model databases" is already pretty crowd, I think
Imustaskforhelp 16 hours ago [-]
This seems really interesting, thanks maxloh for sharing.
I have a quick question but https://aihubmix.com/model/coding-glm-5.1-free seems to be free in the chances of "coding-glm-5.1-free is the open and free version of coding-glm-5.1. To ensure stable service performance, usage limits are in place: up to 5 requests per minute, 500 requests per day, and a daily token allowance of 1 million."
Is there any catch aside from that for this aihubmix? I use opencode-zen from free version mostly if I want agents but this seems interesting to me as well and I think that it mught be able to get integrated into opencode itself as well given this repo is from opencode (well anomalyco)
A quick question but is there any tangible benefit of using these AIhubmix or others over something like opencode-zen itself that I may be missing?
varispeed 13 hours ago [-]
This is missing data like when particular model was nerfed or how often provider routes to cheaper less capable model (variants of so called adaptive reasoning).
Cost per token says nothing. For instance if model goes dumb half way the task and you have to start again. If model does that all the time, then the cost is substantially higher than headline figure.
Probably such a service should constantly run various types of tasks on such models and gauge quality of output (though still provider can detect it and pin their best model to skew the results).
This has been solid and I use this for my Open Source project RightModel: https://rightmodel.dev
nit suggestion: It took a while to realize that I have to scroll to right to see more details. Most users are in habit of scrolling down or click on some button to see more details.
There are literally dozens of existing projects that are doing what you are trying to do.
Insert XKCD standards reference here:
https://www.helicone.ai/llm-cost
https://pricepertoken.com/
https://docs.litellm.ai/docs/provider_registration/add_model...
https://artificialanalysis.ai/api-reference
https://github.com/simonw/llm-prices
https://github.com/assistant-ui/modelpedia
https://github.com/pydantic/genai-prices
https://github.com/Portkey-AI/models
https://github.com/truefoundry/models
https://github.com/agentstation/starmap
https://github.com/dcSpark/ai-model-catalog
https://github.com/mitkury/aimodels
https://github.com/nuxdie/ai-pricing
I always hope for the best when someone has a new list because of this. I want a de facto source!
I was skeptical at first, but I tried using that and then asked Claude if the changes were okay, and they were, and the code works fine.
So I opted for the chinese version. I personally see them as a lesser evil.
I have a quick question but https://aihubmix.com/model/coding-glm-5.1-free seems to be free in the chances of "coding-glm-5.1-free is the open and free version of coding-glm-5.1. To ensure stable service performance, usage limits are in place: up to 5 requests per minute, 500 requests per day, and a daily token allowance of 1 million."
Is there any catch aside from that for this aihubmix? I use opencode-zen from free version mostly if I want agents but this seems interesting to me as well and I think that it mught be able to get integrated into opencode itself as well given this repo is from opencode (well anomalyco)
A quick question but is there any tangible benefit of using these AIhubmix or others over something like opencode-zen itself that I may be missing?
Cost per token says nothing. For instance if model goes dumb half way the task and you have to start again. If model does that all the time, then the cost is substantially higher than headline figure.
Probably such a service should constantly run various types of tasks on such models and gauge quality of output (though still provider can detect it and pin their best model to skew the results).