OmniAI

[OmniAI @OmniAI_1]

@nvandewetering Th point isn’t that it’s bad. It’s that it’s not really a unique product

[OmniAI @OmniAI_1]

@KrishnaBuildsAI Never launched in the first place

[OmniAI @OmniAI_1]

@matthewmillerai They had it going so well with Gemini 2.5 serious all the way to the 3.1 serious.

They kinda got lazy for some reason. Really weird as it feels like they deliberately chose to be bad.

[OmniAI @OmniAI_1]

The only list Gemini 4 argon tops is the text ranking. Beating opus 4.6, but I’m here for the coding personally

[OmniAI @OmniAI_1]

Gemini 4 argon’s rating on LLM arena.

Need I say less? They either took too long to release, or this model is sub par.

I really hope argon is the equivalent of the new flash or something, otherwise it’s not good enough.

[OmniAI @OmniAI_1]

This is their web dev coding rank

[OmniAI @OmniAI_1]

@uzair__afridi @matthias_mrc Very true. Your take is the modern take while Matt’s was the take pre opus 4.5 imo.

[OmniAI @OmniAI_1]

@zakiraicoder Claude

[OmniAI @OmniAI_1]

@bridgemindai Fable 5.5 will probably give more usage then sol lol.

Personally Astra 6.1 will have to make super AGI water to bring me back.

[OmniAI @OmniAI_1]

@Nishant_N_25 Windows can piss of

[OmniAI @OmniAI_1]

@KalCodeDev Is this a better version of bridge mind?

[OmniAI @OmniAI_1]

@IndependentEco I think that fable 5.5 pretty much guaranteed beats astra 6.1 on pure intelligence.

But Argon is the true wild card, we don't know what to expect. Could be really good, even above fable. or mediocre

[OmniAI @OmniAI_1]

If you look at the context of what Tibo is saying. He's not only referring to 6.1 Astra, but to 6.1 Astra being cheaper!

This is because Tibo replied 6.1 coming soon to a post complaining about 6 astra usage.

Perhaps Open AI are still in the race.

[Tibo @thsottiaux]

@davis7 6.1 coming soon

[OmniAI @OmniAI_1]

@ziwenxu_ The OpenAI marketing hits different . Though they have to put their actions where their mou to speaks to make up for dev day.

The release has to be amazing for me to keep my subscription

[OmniAI @OmniAI_1]

Rank these models:

1. Gemini 4 Argon

2. Fable 5.5

3. Astra 6.1

Which do you think will be the best.

[OmniAI @OmniAI_1]

Tibo is probably talking about 6.1 Astra.

This probably will be a great model.

This doesn’t matter if the usage is terrible tho. We are beginning to see a dynamic where expensive models are becoming inaccessible to average people.

Open AI started this trend with usage cut

[Tibo @thsottiaux]

@poteto Hey now. You don’t know what we’re releasing next week

[OmniAI @OmniAI_1]

@shakhzod1001 Impossible to say. Could either be a great investment, or an absolutely horrible one.

I sure hope it’s the former

[OmniAI @OmniAI_1]

@TokenGremlin According to bridge bench itself, the drop isn’t classified as a nerf.

The drop is within normal variance.

[OmniAI @OmniAI_1]

@Da7_Tech Space X probably gonna cook with grok 5 though.

They have to

[OmniAI @OmniAI_1]

@Da7_Tech Grok threw away every advantage they had with 4.7.

4.6 was cheap, fast, and somewhat comparable to the frontier which gave it purpose.

Though now we not only have models which do what 4.6 did but better. We have had no reasonable improvement from grok.

4.7 is garbage