Best GPU for GPT-OSS-20B Locally
Real-time prices and hardware recommendations updated for July 2026.
Adjust Context Length
Slide to popular sizes or type any value manually.
Q3 weights (~8.6GB) plus 32k KV cache reach ~11GB. 12GB cards cover the full context window with headroom.
Q4 weights (~11.5GB) plus 32k KV cache reach ~14GB. 16GB cards are the ideal mainstream setup.
Recommended Hardware
Q8 weights (~23GB) plus 32k KV cache exceed 24GB. 32GB+ cards deliver near-lossless precision without VRAM constraint.
Optimizing Setup for GPT-OSS-20B
Quantization Recommendations
For daily coding and reasoning tasks, Q4_K_M (4-bit quantization) offers the best balance of quality and memory efficiency — it reduces memory requirements by over 70% with minimal quality loss compared to FP16. Q8 and higher presets preserve more fidelity at the cost of significantly higher VRAM usage, which may force layer offloading and hurt throughput.
Recommended Local Software
We recommend using Ollama as the primary runner for local inference due to its automated GPU model splitting and context cache optimizations. For advanced fine-tuning or quantization splits, llama.cpp with Flash Attention compiled natively provides the best granular control.
▸How token speeds are estimated
Two metrics are shown per GPU: Read tok/s (how fast the model ingests your prompt) and Decode tok/s (how fast it streams tokens back). They model fundamentally different bottlenecks.
Read (Prefill)
The prompt is processed in one parallel pass. This is compute-bound: it saturates the GPU's tensor cores.
Decode (Generation)
Each new token requires loading the entire model's active weights from VRAM. This is memory-bandwidth-bound: the GPU stalls waiting for data, not computing.
Weights = (activeParams × bits ÷ 8) × 1.15 overhead. KV cache per step = activeParams × multiplier × contextK.
Architecture utilization factors
Left: decode factor — Right: read factor
Data sources
TFLOPS and memory bandwidth are read from the GPU database. When missing, bandwidth falls back to a hardcoded dictionary.
Limitations
These are analytical estimates, not benchmark results. Use them as a relative comparison, not an absolute performance guarantee.