Fastest
iq3_s
Unknown
50.5 tok/s
Highest token generation speed at 50.5 tok/s.
Hardware Benchmarks
Best Local LLM Performance for RTX 3080 + RTX 4090
Compare RTX 3080 + RTX 4090 benchmark results and see which models actually earn the best completed quality scores.
Hardware Overview
GPU-specific specifications for local LLM planning.
- VRAM
- 36GB
- Architecture
- Multi-GPU
Recommended Models
Models we recommend for RTX 3080 + RTX 4090.
Best quality & Largest context
bartowski/Qwen3.8-27B-GGUF/Qwen3.8-27B-IQ4_XS
27B
42.1 tok/sQuality 90.1Context 256,000 tokens
Highest completed quality score at 90.1. Largest context that fits in 36GB VRAM, at 256,000 tokens.
Frequently asked questions
Which models actually run best on the RTX 3080 + RTX 4090, by task, from community benchmark data.
- What is the best local model for coding on the RTX 3080 + RTX 4090?
- On the RTX 3080 + RTX 4090, bartowski/Qwen3.8-27B-GGUF/Qwen3.8-27B-IQ4_XS ranks first for coding in our benchmarks (89.0/100 at ~42.1 tok/s). iq3_s is faster (~50.5 tok/s) and still scores well (81.6/100) — a good pick if you'd rather trade a little quality for speed.
- What is the best local model for agentic workflows on the RTX 3080 + RTX 4090?
- On the RTX 3080 + RTX 4090, bartowski/Qwen3.8-27B-GGUF/Qwen3.8-27B-IQ4_XS ranks first for agentic workflows in our benchmarks (91.3/100 at ~42.1 tok/s). iq3_s is faster (~50.5 tok/s) and still scores well (86.7/100) — a good pick if you'd rather trade a little quality for speed.
Benchmark Results
Completed quality-ranked benchmark results for NVIDIA GeForce RTX 3080 + NVIDIA GeForce RTX 4090.