Hi team,
I've been impressed by how Atomic-Chat has grown beyond its Jan roots — the
TurboQuant KV cache work, the three-engine architecture behind a single
localhost endpoint, and the recent coding-agent integrations are seriously
good engineering.
Why I'm opening this issue:
I noticed the cloud provider list already includes MiniMax, Qwen, and
Moonshot — clearly the team sees the value of Chinese frontier models. But
the current model means adding each vendor as its own preset with its own
key, and a few notable ones are still missing (DeepSeek, GLM). For users,
each new provider is another account, another key, another billing
relationship to manage.
Suggestion:
Add AnyLLM (https://anyllm.dev) as a single cloud provider preset. It's
an OpenAI-compatible API gateway that puts 24+ models behind one key and
one base URL:
- Chinese frontier models: DeepSeek V4, Qwen3.7-Plus, GLM-5.2 (2M context),
Kimi K2.5/K3, MiniMax M3 (covers your existing presets' flagship models
plus DeepSeek/GLM, through one endpoint)
- International models: GPT-5.6, Gemini 3.5, and more
- Pass-through pricing (zero markup), international cards + crypto accepted
- Built-in routing and automatic failover between providers
This fits Atomic-Chat's "local when you can, cloud when you must" story
nicely: local inference stays private and free, and when users do reach for
the cloud, one preset covers everything — instead of collecting one key per
vendor. Given the app already speaks OpenAI-compatible on both sides
(localhost server + cloud providers), it should be a small preset addition.
Happy to provide API access for testing, or contribute the preset PR if
that's welcome.
(Context: I'm the developer of AnyLLM, an OpenAI-compatible API gateway.)
— Leo Bennett
leo.indiedev@gmail.com
Hi team,
I've been impressed by how Atomic-Chat has grown beyond its Jan roots — the
TurboQuant KV cache work, the three-engine architecture behind a single
localhost endpoint, and the recent coding-agent integrations are seriously
good engineering.
Why I'm opening this issue:
I noticed the cloud provider list already includes MiniMax, Qwen, and
Moonshot — clearly the team sees the value of Chinese frontier models. But
the current model means adding each vendor as its own preset with its own
key, and a few notable ones are still missing (DeepSeek, GLM). For users,
each new provider is another account, another key, another billing
relationship to manage.
Suggestion:
Add AnyLLM (https://anyllm.dev) as a single cloud provider preset. It's
an OpenAI-compatible API gateway that puts 24+ models behind one key and
one base URL:
Kimi K2.5/K3, MiniMax M3 (covers your existing presets' flagship models
plus DeepSeek/GLM, through one endpoint)
This fits Atomic-Chat's "local when you can, cloud when you must" story
nicely: local inference stays private and free, and when users do reach for
the cloud, one preset covers everything — instead of collecting one key per
vendor. Given the app already speaks OpenAI-compatible on both sides
(localhost server + cloud providers), it should be a small preset addition.
Happy to provide API access for testing, or contribute the preset PR if
that's welcome.
(Context: I'm the developer of AnyLLM, an OpenAI-compatible API gateway.)
— Leo Bennett
leo.indiedev@gmail.com