Inference is temporarily unavailable

[farhan-raptas]

Description

Inference is temporarily unavailable for Last 5 minutes on Ubuntu Machine

Model: Deepseek v4 Flash (Opencode Go)

Plugins

No response

OpenCode version

1.17.13

Steps to reproduce

No response

Screenshot and/or share link

No response

Operating System

Ubuntu 24.04

Terminal

No response

[BlackC1over]

same problem in windows
model: DeepSeek v4 flash

[lordrebel]

me too!
my log is here:
model=opencodego:deepseek-v4-flash::session-2026-05-21-b rawModel=deepseek-v4-flash endpoint=chat-completions metadataSource=models.dev messages=5 session=vscode-05834fbd request=req-0e932c15 modelConfiguration={"reasoningEffort":"off"} thinking={"deepseek":"off","glm":"off","kimi":"off","minimax":"off","qwen":"off","qwenBudget":"auto","mimo":"off"} thinkingPayload={}
[request] url=https://opencode.ai/zen/go/v1/chat/completions payloadBytes=122796 requestTimeoutMs=600000 streamIdleTimeoutMs=120000
[http] 503 Service Unavailable content-type=application/json
[http-error-body] {"error":{"message":"Inference is temporarily unavailable","type":"server_error","param":null,"code":"failover_exhausted"}}
[response-summary] status=503 durationMs=26280 ttfbMs=n/a promptTokens=n/a completionTokens=n/a totalTokens=n/a cachedTokens=n/a finishReason= totalBytes=123 totalEvents=0
[2026-07-02T09:20:37.727Z] [go-usage] Recording: provider=OpenCode Go model=deepseek-v4-flash promptTokens=n/a completionTokens=n/a cachedTokens=n/a status=503 error=OpenCode Go API request failed (503) model=deepseek-v4-flash payloadBytes=122796: Inference is temporarily unavailable
[2026-07-02T09:20:37.728Z] [go-usage] After record: entries=0
[2026-07-02T09:20:37.729Z] ERROR model=opencodego:deepseek-v4-flash::session-2026-05-21-b: OpenCode Go API request failed (503) model=deepseek-v4-flash payloadBytes=122796: Inference is temporarily unavailable

but opecode zen can work

[github-actions]

Thanks for updating the issue! It now has the key details needed (model, provider, OS, and version). We appreciate you taking the time to fill things out.

[smok95]

API call failed (attempt 1/3): InternalServerError [HTTP 503]
🔌 Provider: opencode-go Model: deepseek-v4-flash
🌐 Endpoint: https://opencode.ai/zen/go/v1
📝 Error: HTTP 503: Inference is temporarily unavailable
⏱️ Elapsed: 24.89s Context: 2 msgs, ~7,710 tokens
⏳ Retrying in 2.7s (attempt 1/3)...
⚠️ API call failed (attempt 2/3): InternalServerError [HTTP 503]
🔌 Provider: opencode-go Model: deepseek-v4-flash
🌐 Endpoint: https://opencode.ai/zen/go/v1
📝 Error: HTTP 503: Inference is temporarily unavailable
⏱️ Elapsed: 52.48s Context: 2 msgs, ~7,710 tokens
⏳ Retrying in 5.4s (attempt 2/3)...
⚠️ API call failed (attempt 3/3): InternalServerError [HTTP 503]
🔌 Provider: opencode-go Model: deepseek-v4-flash
🌐 Endpoint: https://opencode.ai/zen/go/v1
📝 Error: HTTP 503: Inference is temporarily unavailable
⏱️ Elapsed: 82.80s Context: 2 msgs, ~7,710 tokens
❌ API failed after 3 retries — HTTP 503: Inference is temporarily unavailable
💀 Final error: HTTP 503: Inference is temporarily unavailable

[GGSDDUAY]

me too!

[Johell1NS]

me too

[SamMousa]

Switching to another model supported by opencode go works.

[philiaspaceai]

me too bro, we have the same problem here

[ArnChowdhury]

us bro us

[TaotaoZLG]

+1 bro

[gianghp123]

same for mimo-v2.5

[Raingor]

me too