It's worth trying if you have easy access to it (we use Cursor at work, so it was just a matter of picking a different model for me). Don't know if I'd go out of my way to get a Google AI subscription for it, but if Carbon ends up meeting these expectations, then it may be worth doing so.
It's best to start with a relatively clean GEMINI.md too. Many configs optimized for Claude doesn't work well because Argon didn't have those same failure modes.
It's sad to see Carbon releasing at the same pricing because that implies Google is planning to do the thing everyone is doing: release promising model for subjective preference, RL-maxx a V2, replace promising V1 with V2 because it benchmarks better
Muse did it, Kimi did it, Deepseek did it, Tencent did it. No one seems to care abojt subjective preference anymore.
Top is Opus 5.5, followed by Fable.
- Corrupted/dropped Gemini thought_signature (reasoning_details) in tool loops
- Blinded model.api.id = "contestant" bypasses gemini.txt and serves Claude's prompt (default.txt)
- Anthropic-only prompt caching
Prompt was:
> I read on hacker news the following, which does not match my expectations. Check if something in the harness or bench is hurting gemini perf.
followed by your post.
> anthropic only prompt caching
Anthropic has custom prompt cache requirements, but we've verified all providers are working well.
> Blinded model.api.id = "contestant" bypasses gemini.txt and serves Claude's prompt (default.txt)
This is wrong. We're serving default.txt which is opencode's default prompt. We are not serving a custom prompt for any model, including Gemini.
> Corrupted/dropped Gemini thought_signature (reasoning_details) in tool loops
We blind the reasoning details in the frontend because this would leak information. Some models summarize thinking while others provide raw thinking. I'll check if we're corrupting the thought signature that's actually sent to the backend.
So far this sort of looks like an example of Gemini being bad. Was this Argon?
Their agents leaderboard is literally just randomizing the agent and looking at metadata like tool call errors. In my opinion, it's poor quality data in high volume.
Their web dev leaderboard is in cloud containers, most people don't run agents like that.
I ran a statistical test on our data. Deepseek vs. Argon. Posterior probability DeepSeek 4.1 Flash is stronger: 99.1%.
It's possible, but we are getting real prompts from people on real projects, and the instructions about the battle are pretty minimal.
Argon in low and medium mode had higher quality results than it did when running in high mode, almost every time. My tasks involved the agent finding some details in a raster image and then mark it up with predictions about what it found.
Argon on high would do the work, and then a whole bunch of other stuff i did not ask for. All that extra broke everything else downstream from the analysis.
However, doing the same work with Argon in low effort was on par with results from Haiku 5.5 (xhigh) and GPT-6.1-Terra (medium). So, Argon low works better than Argon high, but for my tasks -- a very weak model relative to how google positions it.
I'm ready for Geimini Unobtanium!
what does "local" coding agents mean? Surely you're not hosting it..
But yes, Opus is a tier above Wonder and Argon with very high confidence.
This might be the first comment I've ever seen saying something nice about the model.
For me it also works the opposite way, when it actually comes out eventually it’ll already be old news.
Having said that, I don't see people excited about using Argon so far, so that might be why Google is announcing this to hype Gemini up a bit.
Anthropic has been doing this with Mythos for quite some time...
I don't understand why HN is so surprised to have an announcement for model existence and a follow-up announcement for model availability. It's really not that unusual, as you say Anthropic has done the same, probably others too (though I can't remember)
it actually did really well.
But it still annoys me: I told it to _tell_ me something, and instead it _implemented_ the something. The previous models were way too eager to just jump in and do and this one seems no different. But the optimization it identified was both really solid and just a good engineering choice, and better than what Opus had been suggesting.
It also gave pretty good feedback on a bit of writing.
Consequently, I have zero intention of renewing my Gemini subscription. What a dumb business choice they made.