Hardware → models
RTX 4060 Ti 16 GB
Start with gpt-oss-20b at MXFP4 profile for general chat. For a coding-focused option, try Qwen2.5-Coder 14B.
Models to start with
Planning estimates16 GB dedicated VRAM · 4,000 tokens · 32 GB system RAM · full GPU residency
| Model / starting file | Why consider it | GPU memory estimate | Memory left¹ |
|---|---|---|---|
| gpt-oss-20bMXFP4 profile | Reasoning and tool-oriented tasksNeeds a runtime that supports MXFP4 and the Harmony conversation format. | 14.8 GB | 0.7 GB |
| Qwen2.5-Coder 14BQ4_K_M | Coding with a larger modelMore memory than the 7B version. Compare them on your own codebase. | 10.2 GB | 5.3 GB |
| Qwen3 8BQ4_K_M | General chat and mixed tasksA middle-sized option with more memory headroom than the 14B models. | 6.2 GB | 9.3 GB |
| Llama 3.2 3B InstructQ4_K_M | Small text assistantA smaller weight footprint. Check the exact Llama checkpoint and licence. | 3.7 GB | 11.8 GB |
From the published model guides here; the Finder covers the full catalogue. ¹ Memory left after the Finder’s reserve. Other applications can use this space. Larger models and higher precision do not guarantee better answers.
gpt-oss-20b fits the short-context planning profile; its MXFP4 runtime requirements still matter. Longer context can exceed 16 GB. CPU offloading changes the memory allocation and may slow generation.
Setup and common questions
Is this the 16 GB version?
NVIDIA lists both 8 GB and 16 GB RTX 4060 Ti variants. Check the exact card’s memory capacity. Extra VRAM gives more room for weights and cache; it does not imply twice the response speed.
How should I try a larger model?
Load one model at 4K context and confirm GPU offload. Check memory during generation. Then change either quantisation or context using the same prompt, thinking mode and number of conversations.
Would a larger GPU help?
Identify whether you need a larger file, longer context or faster responses. A 24 GB RTX 4090 offers more capacity, but test your workload before buying hardware. These comparisons do not establish a measured speed ranking.
How are these recommendations chosen?
We use the main Finder’s balanced task selection and memory calculation, limited to models with published guides here. Task scores and speed proxies are planning inputs, not measured quality rankings.
GPU memory includes weights, context cache and a 0.9 GB runtime allowance, plus a 3% reserve (0.15–1.5 GB). Cache format and runtime buffers can change the fit. See the calculation assumptions.
Sources and related guides
Specification sources checked 2 October 2026.
These comparisons are planning estimates. No reviewed or independently reproduced measurements are available yet. About test evidence.