#Can I set up dreaming mode using gemini api key or openai codex Oauth?
1 messages Β· Page 1 of 1 (latest)
Short version:
- Gemini API key: yes
- OpenAI Codex OAuth: not by itself
Why: dreaming lives under plugins.entries.memory-core.config.dreaming, and the memory stack still needs an embedding provider. The docs explicitly say Codex OAuth covers chat/completions only and does not satisfy embedding requests.
So the practical setup is:
- Gemini API key only β good, easiest path
- Codex OAuth + Gemini API key β also good
- Codex OAuth only β not enough for the embedding side of memory/dreaming
Enable it here:
{
plugins: {
entries: {
"memory-core": {
config: {
dreaming: {
enabled: true,
},
},
},
},
},
}
Then you can use /dreaming on too.
Docs:
- https://docs.openclaw.ai/concepts/dreaming
- https://docs.openclaw.ai/reference/memory-config
- https://docs.openclaw.ai/concepts/model-providers
If you want, paste openclaw --version and openclaw models status, and Iβll tell you the cleanest exact config for your setup.
hello
hey π if you want, I can help you wire up the exact dreaming config for either:
- Gemini-only, or
- Codex OAuth + Gemini embeddings
If you paste openclaw --version and openclaw models status, Iβll keep it precise.
How do I set gemini api key for dreaming mode? on onboarding?
Yep, onboarding is the easiest way.
Use:
openclaw onboard --auth-choice gemini-api-key
That sets up the Google provider auth. Dreaming does not have its own separate API key, it just uses your configured memory/model stack.
A couple details that matter:
- For memory/embeddings, OpenClaw can auto-detect Gemini when
GEMINI_API_KEYis available. - If you want to force it, set
memorySearch.provider = "gemini". - Then enable dreaming with:
{
plugins: {
entries: {
"memory-core": {
config: {
dreaming: {
enabled: true,
},
},
},
},
},
}
Or toggle it in chat with /dreaming on.
One important gotcha: if your Gateway runs as a daemon, make sure GEMINI_API_KEY is available to that process too, for example via ~/.openclaw/.env.
Docs:
- https://docs.openclaw.ai/providers/google
- https://docs.openclaw.ai/concepts/dreaming
- https://docs.openclaw.ai/reference/memory-config
If you want, I can give you the smallest working config for Gemini + dreaming next.
go on
Yep, hereβs the smallest working path.
1. Set the Gemini key during onboarding
openclaw onboard --auth-choice gemini-api-key
That handles the provider auth part.
2. Add the minimal config
If onboarding already stored the key, this is enough:
{
agents: {
defaults: {
memorySearch: {
provider: "gemini",
},
},
},
plugins: {
entries: {
"memory-core": {
config: {
dreaming: {
enabled: true,
},
},
},
},
},
}
3. If you want to wire the key explicitly in config instead
{
agents: {
defaults: {
memorySearch: {
provider: "gemini",
remote: {
apiKey: "${GEMINI_API_KEY}",
},
},
},
},
plugins: {
entries: {
"memory-core": {
config: {
dreaming: {
enabled: true,
},
},
},
},
},
}
4. Then verify
openclaw models status
And if you run the Gateway as a daemon, make sure GEMINI_API_KEY is available to that process too, for example in ~/.openclaw/.env.
Useful docs: