#⚠️ Agent failed before reply: All models failed (1): synthetic/hf:nvidia/Kimi-K2.5-NVFP4: Provider
1 messages · Page 1 of 1 (latest)
That error means OpenClaw tried to use your primary model (synthetic/hf:nvidia/...), but every auth profile for the synthetic provider is currently disabled/cooldown due to billing/credits — so there’s nothing it can use, and because you have no model fallbacks configured (All models failed (1)), the agent dies before it can reply.
A few concrete fixes:
- Confirm + see what it will try next
- Run:
openclaw --version - Run:
openclaw models status(or in Discord:/model status)
- Add at least one fallback model on a different provider
- Run:
openclaw models fallbacks list - Add one:
openclaw models fallbacks add <provider/model>
(Pick from/model listso you don’t guess IDs.)
- Fix the billing issue for
synthetic
- Whatever account/key backs that provider likely hit “insufficient credits / billing disabled”.
- Note: OpenClaw intentionally applies a long backoff for billing failures (starts around 5 hours, can grow up to 24 hours) so it doesn’t hammer a dead-billing provider. Docs: https://docs.openclaw.ai/concepts/model-failover
If you paste openclaw models status output here (redact anything sensitive if it shows up), I can tell you exactly whether you’re missing fallbacks vs missing/disabled auth profiles.
Server@MacBookPro ~ % openclaw models status
🦞 OpenClaw 2026.2.24 (df9a474) — I'll do the boring stuff while you dramatically stare at the logs like it's cinema.
Config : ~/.openclaw/openclaw.json
Agent dir : ~/.openclaw/agents/main/agent
Default : ollama/kimi-k2.5:cloud
Fallbacks (3) : synthetic/hf:nvidia/Kimi-K2.5-NVFP4, synthetic/hf:MiniMaxAI/MiniMax-M2.1, google/gemini-3-flash-preview
Image model : -
Image fallbacks (0): -
Aliases (4) : gemini-flash -> google/gemini-3-flash-preview, gemini -> google/gemini-3-pro-preview, MiniMax M2.1 -> synthetic/hf:MiniMaxAI/MiniMax-M2.1, Kimi-K2.5 NVFP4 -> synthetic/hf:nvidia/Kimi-K2.5-NVFP4
Configured models (4): google/gemini-3-flash-preview, google/gemini-3-pro-preview, synthetic/hf:MiniMaxAI/MiniMax-M2.1, synthetic/hf:nvidia/Kimi-K2.5-NVFP4
Auth overview
Auth store : ~/.openclaw/agents/main/agent/auth-profiles.json
Shell env : off
Providers w/ OAuth/tokens (0): -
- google effective=profiles:~/.openclaw/agents/main/agent/auth-profiles.json | profiles=1 (oauth=0, token=0, api_key=1) | google:default=AIzaSyB6...DOxMyrDw
- ollama effective=models.json:ollama-local | models.json=ollama-local | source=models.json: ~/.openclaw/agents/main/agent/models.json
- synthetic effective=profiles:~/.openclaw/agents/main/agent/auth-profiles.json | profiles=1 (oauth=0, token=0, api_key=1) | synthetic:default=syn_bf68...528e66ff [disabled:billing 3h]
OAuth/token status
- none
Two things jump out from that models status output:
-
You leaked a real Google API key in the thread (
AIzaSy…). Please delete that message and rotate the key in Google Cloud Console ASAP. -
The
synthetickey is explicitly disabled for billing:
synthetic:default=... [disabled:billing 3h]
So OpenClaw is doing the right thing: it’s refusing to use synthetic/* for a while because the provider reported “no credits / billing problem”.
Quick workaround (get replies again)
Switch to your non-synthetic fallback (Gemini) as primary, so it doesn’t even try synthetic first:
CLI
openclaw models set google/gemini-3-flash-preview
openclaw gateway restart
or in Discord
/model gemini-flash(or pickgoogle/gemini-3-flash-preview)
Optionally also reorder/clean fallbacks so you don’t burn time on disabled synthetic entries:
openclaw models fallbacks clear
openclaw models fallbacks add google/gemini-3-flash-preview
# (add synthetic back later once billing is fixed)
Fix the root cause
You need to top up / fix billing for the synthetic provider (or replace that API key). Until then, it’ll keep going back into “disabled:billing” backoff.
If you still see “All models failed (1) …synthetic…”
That usually means the Gateway service is running with different config/auth than your shell. Please paste: