Llama 3 70B (Q8)
Meta Llama 3 70B Instruct, Q8_0 GGUF, batch 1, 4K context.
Primary metric: Tokens / sec (tok/s)
Reference prompts
Representative prompts for this workload. Exact prompts and harness settings still depend on the cited source for each record.
- Prompt 1
A train leaves station A at 9:00 going 60 mph. Another leaves station B (180 miles away) at 9:30 going 80 mph toward A. When do they meet? Show your work.
- Prompt 2
If a 70B-parameter model uses ~140GB at FP16, what is its likely memory footprint at Q4_K_M? Show the arithmetic.
- Prompt 3
Walk through which of these is cheaper for 10M tokens/day at 30 days: H100 cloud at $2/hr vs RTX 4090 owned at $1600 + $0.10/kWh.
- Prompt 4
Explain step-by-step why FP8 inference can be 2x faster than FP16 on H100 but not on A100.
- Prompt 5
Given a 4-bit quantized Llama 3 70B model at ~40GB, how much VRAM headroom is needed for KV cache at 8K context, batch 1?
Reference runtime command
A representative invocation for reproducing this workload class. Source-specific runs may use adjacent runtimes unless the record says otherwise.
llama-server -m llama-3-70b-instruct.Q8_0.gguf -c 4096 -ngl 999 --seed 42Single batch. Q8_0 (~75GB). Often requires multi-GPU or partial CPU offload. Median of 5 runs.
Full leaderboard
Every record for Llama 3 70B Q8.
Top 10 leaderboard
Llama 3 70B (Q8) · sorted by tokens / sec
Value frontier, MSRP vs tokens / sec
Each dot is a device. Top-left is best value (cheap + fast).
Full leaderboard
Click any row for detailed breakdown. Click column headers to sort.
| # | Device | Verif | Buy | |||||||
|---|---|---|---|---|---|---|---|---|---|---|
| 1 | Google TPU v5pASIC Google·INT8·8K ctx | 6 | 110tok/s | 95 GB | 700 W | $38k | 0.16 tok/s/W | $0.0037/k | Amazon | |
| 2 | NVIDIA B200 192GBDatacenter GPU NVIDIA·Q8_0·8K ctx | 6 | 98tok/s | 192 GB | 1000 W | $40k | 0.10 tok/s/W | $0.0043/k | Amazon | |
| 3 | AWS Trainium2ASIC AWS·INT8·8K ctx | 6 | 92tok/s | 96 GB | 500 W | $21k | 0.18 tok/s/W | $0.0024/k | Amazon | |
| #4 | Google TPU v5eASIC Google·INT8·4K ctx | 6 | 62tok/s | 16 GB | 170 W | $9.0k | 0.36 tok/s/W | $0.0015/k | Amazon | |
| #5 | NVIDIA H200 141GBDatacenter GPU NVIDIA·Q8_0·8K ctx | 6 | 58tok/s | 141 GB | 700 W | $30k | 0.08 tok/s/W | $0.0055/k | Amazon | |
| #6 | AMD Instinct MI300X 192GBDatacenter GPU AMD·Q8_0·8K ctx | 6 | 48tok/s | 192 GB | 750 W | $18k | 0.06 tok/s/W | $0.0040/k | Amazon | |
| #7 | NVIDIA H100 SXM5 80GBDatacenter GPU NVIDIA·Q8_0·4K ctx | 6 | 42tok/s | 80 GB | 700 W | $25k | 0.06 tok/s/W | $0.0063/k | Amazon | |
| #8 | AMD Instinct MI300X 192GBDatacenter GPU AMD·Q8_0·8K ctx | 6 | 38tok/s | 192 GB | 750 W | $15k | 0.05 tok/s/W | $0.0042/k | Amazon | |
| #9 | NVIDIA DGX Spark (Project DIGITS, 128GB)Datacenter GPU NVIDIA·Q8_0·8K ctx | 6 | 22tok/s | 128 GB | 240 W | $3.0k | 0.09 tok/s/W | $0.0014/k | Amazon | |
| #10 | 2× NVIDIA RTX 4090 24GBConsumer GPU NVIDIA·Q8_0·4K ctx | 6 | 18tok/s | 48 GB | 900 W | $3.2k | 0.02 tok/s/W | $0.0019/k | Amazon | |
| #11 | 4× NVIDIA RTX 3090 24GBConsumer GPU NVIDIA·Q8_0·4K ctx | 6 | 14tok/s | 96 GB | 1400 W | $4.0k | 0.01 tok/s/W | $0.0030/k | Amazon | |
| #12 | NVIDIA GeForce RTX 5090 32GBConsumer GPU NVIDIA·Q8_0·4K ctx | 6 | 14tok/s | 32 GB | 575 W | $2.0k | 0.02 tok/s/W | $0.0015/k | Amazon | |
| #13 | Apple M3 Ultra (80c GPU, 512GB)Apple Silicon Apple·Q8_0·8K ctx | 6 | 9tok/s | 512 GB | 270 W | $9.5k | 0.03 tok/s/W | $0.0112/k | Amazon | |
| #14 | Apple M4 Max (40c GPU, 128GB)Apple Silicon Apple·Q8_0·4K ctx | 6 | 5tok/s | 128 GB | 65 W | $5.0k | 0.08 tok/s/W | $0.0102/k | Amazon |
Cite this benchmark
Use this in your paper, blog post, or comparison table.
@misc{myaihardware_llama3-70b-q8_2026,
title = {MyAI Bench: Llama 3 70B (Q8)},
author = {{MyAIHardware Contributors}},
year = {2026},
url = {https://www.myaihardware.com/benchmarks/workload/llama3-70b-q8},
note = {Version 1.3, accessed 2026-08-27}
}MyAIHardware Contributors. (2026). MyAI Bench: Llama 3 70B (Q8). MyAIHardware. Retrieved 2026-08-27, from https://www.myaihardware.com/benchmarks/workload/llama3-70b-q8
MyAIHardware Contributors. "MyAI Bench: Llama 3 70B (Q8)." MyAIHardware, 2026, https://www.myaihardware.com/benchmarks/workload/llama3-70b-q8. Accessed 2026-08-27.
MyAI Bench, Llama 3 70B (Q8). MyAIHardware Contributors, 2026. Version 1.3. https://www.myaihardware.com/benchmarks/workload/llama3-70b-q8 (accessed 2026-08-27).