Hermes Mixture of Agents DESTROYS Claude Fable 5? artwork

Hermes Mixture of Agents DESTROYS Claude Fable 5?

AI News Today | Julian Goldie Podcast

July 2, 2026

Fable 5 vs Hermes Mixture of Agents vs Fusion: Side-by-Side Build Tests (Games, WebOS, Landing Pages)The video compares Anthropic’s Fable 5, Hermes Mixture of Agents, and Open Router’s Fusion by testing 42 builds and showing side-by-side results across multiple projects, including a racer game,...
Speakers: Julian Goldie
**Julian Goldie** (0:00)
So Fable 5 versus Hermes Mixture of Agents versus Open Router's Fusion, who wins? Today, we've tested them out. I'm gonna show you the results in a second. This is something we actually built with Fable 5 a few minutes ago. So let's get straight into this. We've compared them side by side. I'm gonna show you exactly what the difference is. Now, if you're wondering, okay, what is Hermes Mixture of Agents? This is a way of combining multiple models together to get better outputs than you'd normally get with, for example, something that's frontier. So, for example, if you wanted to create outputs that are better than Opus 4.8, you could combine GPT 5.5 and Opus 4.8 and that tends to work better. And that's something that I've tested personally, it's also on the benchmarks too. Now, Fusion works in a similar way. So with Fusion, you basically have multiple agents working together as a panel, as you can see right here, then a judge fuses all the answers and they go into one answer. And of course, Anthropic's Fable 5 just came back again. We've tested it relentlessly with 42 different builds, which I'll break down in a second and you can see how they perform side by side. So let's get straight into this. One of the first things we did was test out a game side by side. So let's have a look at this. So this is the example from Hermes Mixture of Agents. This is from Fable 5 and this is from Fusion. Now, out of all the tests, we have a look side by side. I would say that Fable 5 actually created the best output so far. And bear in mind, the other tests with Fusion and Hermes Mixture of Agents were using non-Fable 5 models. So for example, GPT 5.5 with Opus 4.8. And it is created the nicest output as you can see right here. Next up, we have the Nordic Crypt game. So this is from Fusion.
Super dark, pretty hard to see what's going on. It kind of works, but doesn't work at the same time. Like you wouldn't play that game at all. If we have a look at the output from Hermes Mixture of Agents, quite nice, works nicely, little bit hard to control, but it does work well. I also like the moving background as well. It looks quite nice. Good vibe to it.
Now if we have a look at the one from Fable 5, I would say this is the number one, the easiest to navigate. Number two, it's got a lot more to it than the other two. Like the other two seem to be a bit more basic. This has more detail as well. It's easier to navigate and move around. So for the crypt game as well, I would say that Fable 5 is winning right here. So the interesting thing about this is if you look at Hermes Mixture of Agents, if you look at particularly Fusion, which Fusion claims to beat Fable 5 on benchmarks, that's not what I'm seeing so far. But there's many more tests to try out. Then we've got the orbit test from our agents too. So this is with Fusion, not bad at all. It's in a solar system orbit. So it's like a diagram of the planets and everything else. You can change the speed and the controls over here.
Then we've got the version from Hermes Mixture of Agents. You can also drag and move items into the galaxy as well, which is pretty nice. You can change the speed over here. The controls are good. You can reset it. You can pause it as well. Some good controls on that. Now if we have a look at Fable 5's output, I would say this is the most basic. So it's not always the case that Fable 5 wins on every single test. As you can see on this one, it didn't do so well. I would say Hermes Mixture of Agents won that. So we have three tests in. Fable 5 has won two of them. Hermes has won one of them. Fusion hasn't won any yet. Next up we have the landing page test. This is with Fable 5 And this is actually something I've noticed with Fable 5, is like it kind of struggles with UI. So if you look at this particular output, it doesn't look so nice. It's a little bit messy. The background is moving, but kind of feels weird. Everything's just a bit off in this particular example. Now, if we have a look at Nova, this is the output from Fusion.
It's definitely beating Fable 5 So it definitely outperforms Fable 5 on landing page creation for this particular test. Now, if we have a look at Hermes Mixture of Agents, I would say this is by far the nicest output that we got. See how it's glowing? When I move the mouse, the background is moving, but still looks quite nice.

9 more minutes of transcript below

Feed this to your agent

Try it now — copy, paste, done:

curl -H "x-api-key: pt_demo" \
  https://spoken.md/transcripts/1000651996090

Works with Claude, ChatGPT, Cursor, and any agent that makes HTTP calls.

From $0.10 per transcript. No subscription. Credits never expire.

Using your own key:

curl -H "x-api-key: YOUR_KEY" \
  https://spoken.md/transcripts/1000775248429