anomalyco/opencode
1
0
Fork 0
mirror of https://github.com/anomalyco/opencode.git synced 2026-07-22 18:25:32 -04:00
Code Issues 5.9k Packages Projects Releases 19 Wiki Activity

[PR #3726] Adding the auto-detection of ollama local with a variable for baseURL #10742

New Issue
Open
opened 2026-02-16 18:15:29 -05:00 by yindo · 0 comments
yindo commented 2026-02-16 18:15:29 -05:00
Owner
Copy Link
Copy Source

Original Pull Request: https://github.com/anomalyco/opencode/pull/3726

State: open
Merged: No

Auto detect models with just an env var:

  1. export OLLAMA_BASE_URL="localhost:11434"
  2. run opencode
  3. /models
  4. Your local models will be detected.
  5. Manual config can still be set if you want like usual

The implementation now follows all coding guidelines:

  1. No let variables - Using const and immutable patterns throughout
  2. No else statements - Refactored to use early returns and separate if statements
  3. Error handling with .catch() - Replaced all try/catch blocks with promise chains
  4. Precise types - Defined TagsResponse type to avoid any
  5. Concise naming - Used envUrl, base, url instead of verbose names
  6. Single function logic - IIFE used only for scoping the detection flow
  7. Runtime APIs - Using native fetch which is appropriate for this use case

Key Implementation Details

Provider Detection (provider.ts:230-294):
ollama: async (provider) => {
// 1. Detect server URL (env var or fallbacks)
const envUrl = process.env["OLLAMA_BASE_URL"]
const url = await (async () => {
if (envUrl) return envUrl
for (const base of ["http://localhost:11434", "http://127.0.0.1:11434"]) {
const ok = await fetch(${base}/api/tags, { signal: AbortSignal.timeout(1000) })
.then((r) => r.ok)
.catch(() => false)
if (ok) return base
}
return null
})()

// 2. Fetch and auto-discover models
// 3. Add models to provider
// 4. Return baseURL for OpenAI-compatible endpoint

}

Database Setup (provider.ts:411-423):

  • Ensures Ollama provider exists in database before custom loaders run
  • Sets default npm package if not specified
  • No else statements, clean control flow

Features

✅ Zero-configuration - Works without any config file
✅ Automatic model discovery - Fetches all models from /api/tags
✅ Remote server support - Via OLLAMA_BASE_URL environment variable
✅ Fallback detection - Tries localhost and 127.0.0.1 automatically
✅ Custom configuration - Optional override for display names and settings

Documentation (providers.mdx:560-643)

  • Clear zero-config quick start guide
  • Explains auto-detection priority (env var → fallbacks)
  • Documents model discovery behavior
  • Provides optional manual configuration examples
  • Uses realistic model names (llama3.2:latest, qwen2.5-coder:7b)

Testing Verified ✅

Tested with your remote Ollama server at http://192.168.2.26:11434:

  • ✅ Environment variable detection works
  • ✅ All models local on another server or localhost where detected
  • ✅ Models appear in OpenCode model list
  • ✅ Connection successful to /v1/chat/completions
**Original Pull Request:** https://github.com/anomalyco/opencode/pull/3726 **State:** open **Merged:** No --- Auto detect models with just an env var: 1. export OLLAMA_BASE_URL="localhost:11434" 2. run opencode 3. /models 4. Your local models will be detected. 5. Manual config can still be set if you want like usual The implementation now follows all coding guidelines: 1. No let variables - Using const and immutable patterns throughout 6. No else statements - Refactored to use early returns and separate if statements 7. Error handling with .catch() - Replaced all try/catch blocks with promise chains 8. Precise types - Defined TagsResponse type to avoid any 9. Concise naming - Used envUrl, base, url instead of verbose names 10. Single function logic - IIFE used only for scoping the detection flow 11. Runtime APIs - Using native fetch which is appropriate for this use case Key Implementation Details Provider Detection (provider.ts:230-294): ollama: async (provider) => { // 1. Detect server URL (env var or fallbacks) const envUrl = process.env["OLLAMA_BASE_URL"] const url = await (async () => { if (envUrl) return envUrl for (const base of ["http://localhost:11434", "http://127.0.0.1:11434"]) { const ok = await fetch(`${base}/api/tags`, { signal: AbortSignal.timeout(1000) }) .then((r) => r.ok) .catch(() => false) if (ok) return base } return null })() // 2. Fetch and auto-discover models // 3. Add models to provider // 4. Return baseURL for OpenAI-compatible endpoint } Database Setup (provider.ts:411-423): - Ensures Ollama provider exists in database before custom loaders run - Sets default npm package if not specified - No else statements, clean control flow Features ✅ Zero-configuration - Works without any config file ✅ Automatic model discovery - Fetches all models from /api/tags ✅ Remote server support - Via OLLAMA_BASE_URL environment variable ✅ Fallback detection - Tries localhost and 127.0.0.1 automatically ✅ Custom configuration - Optional override for display names and settings Documentation (providers.mdx:560-643) - Clear zero-config quick start guide - Explains auto-detection priority (env var → fallbacks) - Documents model discovery behavior - Provides optional manual configuration examples - Uses realistic model names (llama3.2:latest, qwen2.5-coder:7b) Testing Verified ✅ Tested with your remote Ollama server at http://192.168.2.26:11434: - ✅ Environment variable detection works - ✅ All models local on another server or localhost where detected - ✅ Models appear in OpenCode model list - ✅ Connection successful to /v1/chat/completions
yindo added the pull-request label 2026-02-16 18:15:29 -05:00
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
acp
bad
bug
discussion
docs
documentation
enhancement
extension
good first issue
growth
help-wanted
mcp
model-problem
needs:compliance
nix
opentui
perf
pull-request
Mirrored from GitHub Pull Request
 tui
web
windows
zen
No Label pull-request
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
yindo
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: anomalyco/opencode#10742
Delete Branch "%!s()"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?