4 LLMs Tested in Codex, Claude Code, Hermes & OpenClaw (FinAI)
Unlock all features
FREE: Get instant access to 10 AI summaries, chats, or transcripts per day.
Unlock all features
FREE: Get instant access to 10 AI summaries, chats, or transcripts per day.
Unlock all features
FREE: Get instant access to 10 AI summaries, chats, or transcripts per day.
Unlock all features
FREE: Get instant access to 10 AI summaries, chats, or transcripts per day.
Unlock all features
FREE: Get instant access to 10 AI summaries, chats, or transcripts per day.
Related videos
We Finally Know Why T-Rex Had Those Tiny Arms + Other Discoveries
Anton Petrov
297.6k views
Claude Opus 4.8 Is Too Smart… and TOO HONEST
Wes Roth
53.4k views
I'm switching to Hermes (goodbye OpenClaw!!)
NetworkChuck
248.4k views
Lutnick Under Fire on EPSTEIN in CLOSED DOOR Testimony
Legal AF
55.3k views
OpenAI Misses Targets, Codex vs Claude, Elon vs Sam Trial, Big Hyperscaler Beats, Peptide Craze
All-In Podcast
67.6k views
HERMES AGENT SETUP: the OpenClaw killer is here
Wes Roth
25.5k views
Claude just became OpenClaw
Wes Roth
37.3k views
Hegseth Freezes ... Incriminates Trump In Air Force One Disaster
Jack Cocchiarella
256.3k views
Head of Claude Code: What happens after coding is solved | Boris Cherny
Lenny's Podcast
29.2k views
AI Discovers Anomalies in Hubble Images We Never Knew Existed
Anton Petrov
45.9k views
Top Comments (5)
I have used both OpenClaw and Hermes and side by side. One thing I noticed is OpenClaw would "win" on first try of more things while Hermes would "learn" and "win" if you (a) tell it to learn from what happened; and/or (b) do the task enough times for Hermes to learn on its own. I found after 2 weeks of use of both of them that Hermes pulled far ahead with "reasoning", task completion, token usage, and speed of execution. It constantly refined itself. Eventually, the differences were so stark that I uninstalled OpenClaw.
I couldn't possibly keep up with the important breakthroughs and analysis' of reasoning NLP s and agentic workflow systems . Thanks goodness I rely on rigorous mathematical standards provided on this channel. Thanks in advance for the next period until I comment again.
Hey the paper specified on the left is wrong, it has to be 2605.14355
I think Qwen probably performs best with Hermes because most of the Hermes devs use Qwen locally but Grok now officially support Hermes so maybe the Grok will be better supported to.
Why the fuck did the researches test Qwen3.5 instead of Qwen3.6 27B ... Qwen3.5 fails on most tool calls and much more....
Unlock the Data Inside
Turn Videos into Knowledge
- Get FREE 10/day: transcripts, summaries, chats
- Chat with videos, export text & PDF
- $1 free API credit for RAG, chatbots & research
Free forever plan • All features unlocked
Top Comments (5)
I have used both OpenClaw and Hermes and side by side. One thing I noticed is OpenClaw would "win" on first try of more things while Hermes would "learn" and "win" if you (a) tell it to learn from what happened; and/or (b) do the task enough times for Hermes to learn on its own. I found after 2 weeks of use of both of them that Hermes pulled far ahead with "reasoning", task completion, token usage, and speed of execution. It constantly refined itself. Eventually, the differences were so stark that I uninstalled OpenClaw.
I couldn't possibly keep up with the important breakthroughs and analysis' of reasoning NLP s and agentic workflow systems . Thanks goodness I rely on rigorous mathematical standards provided on this channel. Thanks in advance for the next period until I comment again.
Hey the paper specified on the left is wrong, it has to be 2605.14355
I think Qwen probably performs best with Hermes because most of the Hermes devs use Qwen locally but Grok now officially support Hermes so maybe the Grok will be better supported to.
Why the fuck did the researches test Qwen3.5 instead of Qwen3.6 27B ... Qwen3.5 fails on most tool calls and much more....