⌘K

AI GPU Benchmark Results

Every benchmark number in our database, with its source and evidence type. Filter by workload or evidence type; sort by any column. Independent benchmarks come from third-party reviews and standardized databases (MLPerf); community results are labeled reproducible when the run config is published, estimated when approximated.

100sourced results
27products covered
66GPUs in database
Honest coverage: our current benchmark set covers LLM inference (tokens/s) and Stable Diffusion image generation on flagship and mid-range cards. Gaming FPS and video-generation numbers are not yet sourced — they will appear here when we have verifiable sources. See the methodology and evidence-type definitions.

Filter results

All LLM inference (tokens/s) Image generation Video generation Speech & transcription Fine-tuning
All Manufacturer benchmark Independent benchmark Reproducible community benchmark Unverified community result Estimated
GeForce RTX 4090 view all GPUs

100 benchmark results

GPU Workload Model / config Result Evidence Source
GeForce RTX 4090 LLM inference (tokens/s) Llama-2-13B · UL Procyon AI Text Generation (TensorRT) (Standardized UL Procyon suite; same lab platform (ThreadRipper 7980X, driver 571.86); output rate 92.853 tok/s) 5,013 score Independent benchmark0.85 StorageReview GPU Reviews
GeForce RTX 4090 LLM inference (tokens/s) Llama-3-70B · Q4 · llama.cpp (24GB VRAM, very tight, heavy swapping — editorial estimate pending article-level source verification) 18 tok/s EST Estimated0.40 Tom's Hardware GPU Benchmarks 2025
GeForce RTX 4090 LLM inference (tokens/s) Llama-3-8B · UL Procyon AI Text Generation (TensorRT) (Standardized UL Procyon suite; same lab platform (ThreadRipper 7980X, driver 571.86); output rate 150.039 tok/s) 4,849 score Independent benchmark0.85 StorageReview GPU Reviews
GeForce RTX 4090 LLM inference (tokens/s) Llama-3.1-8B · FP16 · vLLM (Batch-serving throughput, ~18GB VRAM used; published llama.cpp/vLLM benchmarks (approximate)) 2,550 tok/s EST Estimated0.40 Spheron GPU Benchmark Blog
GeForce RTX 4090 LLM inference (tokens/s) Llama-3.1-8B · Q4_K_M · llama.cpp b3500 (CUDA) (Token generation: single user, batch 1, 4k context, mean of 3 runs) 125 tok/s Unverified community result0.55 MyAIHardware llama.cpp Benchmarks
GeForce RTX 4090 LLM inference (tokens/s) Llama-3.1-8B · Q4_K_M · llama.cpp b3500 (CUDA) (Prompt processing: 512-token prompt, batch 1, mean of 3 runs) 4,800 tok/s Unverified community result0.55 MyAIHardware llama.cpp Benchmarks
GeForce RTX 4090 LLM inference (tokens/s) Mistral-7B · UL Procyon AI Text Generation (TensorRT) (Standardized UL Procyon suite; same lab platform (ThreadRipper 7980X, driver 571.86); output rate 183.266 tok/s) 5,094 score Independent benchmark0.85 StorageReview GPU Reviews
GeForce RTX 4090 LLM inference (tokens/s) Phi-3.5-mini · UL Procyon AI Text Generation (TensorRT) (Standardized UL Procyon suite; same lab platform (ThreadRipper 7980X, driver 571.86); output rate 244.343 tok/s) 4,958 score Independent benchmark0.85 StorageReview GPU Reviews
GeForce RTX 4090 LLM inference (tokens/s) Qwen3-8B · Q4_K_XL · llama.cpp llama-bench, CUDA 12.8 (Token generation: 16K context, batch 1, Ubuntu 24.04, on-hardware lab measurement) 104.3 tok/s Unverified community result0.55 Hardware Corner LLM GPU Rankings

How these numbers are sourced

Every row links to its original source and carries an evidence type plus a confidence score. Independent benchmarks come from third-party reviews (Tom's Hardware, TechPowerUp, Puget Systems) and standardized databases (MLPerf); manufacturer benchmarks are vendor-published figures; community results are reproducible when the run config is published, unverified otherwise — always labeled EST when estimated. Read the full benchmark methodology.

Compare any two GPUs head-to-head in the comparison tool, or browse value rankings.