alt.hn

1/22/2026 at 5:26:12 PM

Show HN: First Claude Code client for Ollama local models

 https://github.com/21st-dev/1code

by SerafimKorablev

1/22/2026 at 9:24:25 PM

The Anthropic API was already supported by llama.cpp (The project Ollama ripped off and typically lags in features by 3-6 months), which already works perfectly fine with Claude Code by setting a simple environment variable.

by oceanplexian

1/25/2026 at 3:35:31 PM

Point of clarification: llama.cpp is MIT-licensed. Using it downstream (commercially or otherwise) is exactly what that license allows, so calling it a rip-off is misleading.

by davely

1/22/2026 at 9:37:23 PM

And they reference that announcement and related information in the second line.

by xd1936

1/22/2026 at 10:45:03 PM

Which announcement are you looking at? I see no references to llama-cpp in either Ollama's blog post or this project's github page.

by gcr

1/23/2026 at 12:47:46 PM

As others said this was possible for months already with llama-cop’s support for Anthropic messages API. You just need to set the ANTHROPIC_BASE_URL. The specific llama-server settings/flags were a pain to figure out and required some hunting, so I collected them in this guide to using CC with local models:

https://github.com/pchalasani/claude-code-tools/blob/main/do...

One tricky thing that took me a whole day to figure out is that using Claude Code in this setup was causing total network failures due to telemetry pings, so I had to set this env var to 1: CLAUDE_CODE_DISABLE_NONESSENTIAL_TRAFFIC

by d4rkp4ttern

1/22/2026 at 9:19:02 PM

There are already various proxies to translate between OpenAI-style models (local or otherwise) and an Anthropic endpoint that Claude Code can talk to. Is the advantage here just one less piece of infrastructure to worry about?

by eli

1/22/2026 at 9:54:06 PM

siderailing here - but got one that _actually_ works?

in particular i´d like to call claude-models - in openai-schema hosted by a reseller - with some proxy that offers anthropic format to my claude --- but it seems like nothing gets to fully line things up (double-translated tool names for example)

reseller is abacus.ai - tried BerriAI/litellm, musistudio/claude-code-router, ziozzang/claude2openai-proxy, 1rgs/claude-code-proxy, fuergaosi233/claude-code-proxy,

by g4cg54g54

1/22/2026 at 11:28:59 PM

What probably needs to exist is something like `llsed`.

The invocation would be like this

    llsed --host 0.0.0.0 --port 8080 --map_file claude_to_openai.json --server https://openrouter.ai/api
Where the json has something like

    { tag: ... from: ..., to: ..., params: ..., pre: ..., post: ...}
So if one call is two, you can call multiple in the pre or post or rearrange things accordingly.

This sounds like the proper separation of concerns here... probably

The pre/post should probably be json-rpc that get lazy loaded.

Writing that now. Let's do this: https://github.com/day50-dev/llsed

by kristopolous

1/23/2026 at 4:56:56 AM

Some unsolicited advice: Streaming support is tricky. I'd strip the streaming out when you proxy until everything else is solid.

by eli

1/23/2026 at 2:37:06 PM

Cool. Sounds good. Thanks. I'll do it.

This will be a bit challenging I'm sure but I agree, litellm and friends do too many things and take too long to get simple asks from

I've been pitching this suite I'm building as "GNU coreutils for the LLM era"

It's not sticking and nobody is hyped by it.

I don't know if I should keep going or if this is my same old pattern cropping up again of things I really really like but just kinda me

by kristopolous

1/24/2026 at 4:46:06 PM

So I've pitched this a few more times. It's way too complicated for people.

The value comprehension market is small

So I'll need to surface it better or just do something else

by kristopolous

1/23/2026 at 4:53:06 AM

I've been hacking on this one for a few months now and it works for me https://github.com/elidickinson/claude-code-mux Been optimizing for routing to different models within one session so maybe overkill.

But I'm surprised litellm (and its wrappers) don't work for you and I wonder if there's something wrong with your provider or model. Which model were you using?

by eli

1/22/2026 at 11:13:21 PM

What hardware are you running the 30b model on? I guess it needs at least 24GB VRAM for decent inference speeds.

by dsrtslnd23

1/27/2026 at 1:10:03 AM

Im running qwen3-coder:30b-a3b-q8_0 @ 32k context. Comes out to 36gb and Im splitting it between a 3090 24gb and a 4060ti 16gb (ollama put 20gb on the 3090 and 13.5 on the 4060ti) , runs great tbh. Ollama running in ubuntu server and Im running claude code from my windows desktop pc.

by derp-mcgee

1/23/2026 at 12:54:09 AM

The general rule to follow is that you need as much VRAM as the model size. 30b models are usually around 19GB. So, most likely a GPU with 24GB of VRAM.

by thtmnisamnstr

1/23/2026 at 9:50:03 AM

But this also means tiny context windows. You can't fit gpt-oss:20b + more than a tiny file + instructions into 24GB

by 3836293648

1/24/2026 at 6:37:03 AM

Gpt-oss is natively 4-bit, so you kinda can

by blizdiddy

1/25/2026 at 7:04:36 PM

You can fit the weights + a tiny context window into 24GB, absolutely. But you can't fit anything of any reasonable size. Or Ollama's implementation is broken, but it needs to be restricted beyond usability for it not to freeze up the entire machine when I last tried to use it.

by 3836293648

1/22/2026 at 11:35:48 PM

I'd like to know this, too. I'm just getting started getting my feet wet with ollama and local models using just CPU, and it's obviously terribly slow (even 24 cores, 128GB DRAM. It's hard to gauge how much GPU money I'd need to plonk down to get acceptable performance for coding workflows.

by ryandrake

1/23/2026 at 7:38:54 AM

I tried to build a similar local stack recently to save on API costs. In practice I found the hardware savings are a bit of a mirage for coding workflows. The local models hallucinate just enough that you end up spending more in lost time debugging than you would have paid for Sonnet or Opus to get it right the first time.

by storystarling

1/22/2026 at 10:47:19 PM

I was trying to get Claude code to work with llama.cpp but could never figure out anything functional. It always insisted on a phone home login for first time setup. In cline I’m getting better results with glm-4.7-flash than with qwen3-coder:30b

by horacemorace

1/22/2026 at 11:20:40 PM

~/.claude.json with {"hasCompletedOnboarding":true} is the key, then ANTHROPIC_BASE_URL and ANTHROPIC_AUTH_TOKEN work as expected

by g4cg54g54

1/23/2026 at 12:42:35 PM

Curious what llama-server flags you used. On my M1 Max 64GB MacBook I tried it in Claude Code (which has a 25K system message) and I get 3 tps.

But with Qwen3-30B-A3B I get 20 tps in CC.

by d4rkp4ttern

1/22/2026 at 9:03:17 PM

this is cool. not sure it is the first claude code style coding agent that runs against Ollama models though. goose, opencode and others have been able to do that a while no?

by dosinga

1/22/2026 at 9:42:49 PM

Does this UI work with Open Code?

by d0100

1/22/2026 at 11:05:54 PM

[dead]

by smissingham