Fastest
ornith-1.5-9b-obliterated
9B
89.3 tok/s
Highest token generation speed at 89.3 tok/s.
Hardware Benchmarks
Best Local LLM Performance for 2× RTX 4070
Compare 2× RTX 4070 benchmark results and see which models actually earn the best completed quality scores.
Hardware Overview
GPU-specific specifications for local LLM planning.
- VRAM
- 24GB
- Architecture
- Multi-GPU
Recommended Models
Models we recommend for 2× RTX 4070.
Best quality & Largest context
Ternary-Bonsai-2-27B-Abliterated-v2-PQ2_0-MTP
27B
56.9 tok/sQuality 82.5Context 65,536 tokens
Highest completed quality score at 82.5. Largest context that fits in 24GB VRAM, at 65,536 tokens.
Models on the 2× RTX 4070
Every model the community ran on this hardware. Up is better answers, right is faster. Time to finish counts every token the model wrote, thinking included, so a model that talks a lot ends up further left.
Best trade-off: quality 60+, and nothing else here is both faster and betterOther models · dot size = number of runs
Medians over 7 fully rated runs of 5 models. Each quantization is its own model.
Frequently asked questions
Which models actually run best on the 2× RTX 4070, by task, from community benchmark data.
- What is the best local model for coding on the 2× RTX 4070?
- On the 2× RTX 4070, Ternary-Bonsai-2-27B-Abliterated-v2-PQ2_0-MTP ranks first for coding in our benchmarks (62.8/100 at ~54.7 tok/s).
- What is the best local model for agentic workflows on the 2× RTX 4070?
- On the 2× RTX 4070, Swift-Bonsai-2-PQ2_0 ranks first for agentic workflows in our benchmarks (90.7/100 at ~44.1 tok/s).
Benchmark Results
Completed quality-ranked benchmark results for 2× NVIDIA GeForce RTX 4070.