Fastmode=true gpt 5.4 | Friends of the Crustacean 🦞🤝 | Page 1

verbal bramble Apr 12, 2026, 8:35 PM

#

Excuse me guys but how can we enable fast mode in gpt 5.4 inside openclaw using gpt 5.4 codex subscription

primal pivotBOT Apr 12, 2026, 8:35 PM

#

glass geyserBOT Apr 12, 2026, 8:35 PM

#

To help others find answers, you can mark your question as solved via Right click solution message -> Apps -> ✅ Mark Solution

primal pivotBOT Apr 12, 2026, 8:35 PM

#

eternal elbowBOT Apr 12, 2026, 8:36 PM

#

Yep, but with one important catch:

Your Codex/ChatGPT subscription maps to openai-codex/gpt-5.4, not openai/gpt-5.4.

So the usual flow is:

switch to the Codex OAuth model
```
/model openai-codex/gpt-5.4
```
enable fast mode
```
/fast on
```

If you want it on by default, set:

#

{
  agents: {
    defaults: {
      models: {
        "openai-codex/gpt-5.4": {
          params: {
            fastMode: true,
          },
        },
      },
    },
  },
}

If you're trying to use openai/gpt-5.4, that needs a direct OpenAI API key, not the Codex subscription.

#

Docs:

If you want, paste your openclaw --version and openclaw models status, and I can sanity-check the exact setup.

verbal bramble Apr 13, 2026, 4:39 PM

#

What do fast mode exactly do

eternal elbowBOT Apr 13, 2026, 4:40 PM

#

For GPT-5.4 in OpenClaw, fast mode is basically a priority-processing flag.

With openai-codex/gpt-5.4, /fast on makes OpenClaw send:

service_tier=priority

So, practically:

It can reduce latency / make replies start faster
It does not switch to a different model
It does not increase intelligence/quality by itself
It does not change reasoning or verbosity settings
If you're going through a proxy instead of native OpenAI/Codex endpoints, it may do nothing

So I’d think of /fast as “please use the faster priority lane if this provider/account supports it”, not “turbo mode” in the model itself.

#Fastmode=true gpt 5.4