8GB GPU offline pair programmer (Continue.dev + Ollama)
A tested guide (Ollama v0.24.0, Continue v1.2.22) for running chat, tab-completion and codebase search on an 8GB NVIDIA GPU.
SoloCodingUnder $1kSensitivity: High (client/confidential)
Hardware
| Item | Approx. price (as of 2026-10-16) |
|---|---|
| On sale −6% MSI GeForce RTX 5060 Ti 16GB Ventus 3X OC (Newegg, related current offer) | $579.99 |
| NVIDIA GeForce RTX 4060 (8GB) or similar 8GB GPU | Not stated in source |
Source prices are as stated by the source. Retailer offers: prices checked 2026-10-16, may change. Some links may be affiliate links; we may earn a commission at no extra cost to you. This never changes what we list.
Software stack
- Ollama
- Continue.dev (VS Code)
Models used
- qwen2.5-coder:7b (q4_K_M, ~4.7GB VRAM) for chat/edit
- qwen2.5-coder:1.5b for autocomplete
- nomic-embed-text for @Codebase
Good at
- Sub-400ms tab completion on an 8GB GPU
- @Codebase indexing for project context
- ~15 minute setup
Weak at
- CPU-only is much slower: 20-40s per chat response vs 3-6s on GPU
Rough total cost
Not stated in source (assumes an existing PC with an 8GB GPU).
Getting started
- Install Ollama.
- ollama pull qwen2.5-coder:7b && ollama pull qwen2.5-coder:1.5b && ollama pull nomic-embed-text
- Install Continue in VS Code.
- Add the three models to config.yaml with roles chat/edit/apply, autocomplete (debounceDelay 300, maxPromptTokens 400) and embed.
- Test in the sidebar and with @Codebase.
Source & credit
Continue.dev + Ollama Setup Guide: Pair Programming Offline — aifoss.dev
Summarized from the source; quotes are the source author's words. Check the original before buying.