Hardware › NVIDIA RTX 5070 Ti 16GB
NVIDIA RTX 5070 Ti 16GB for local AI: price in Australia & what it can run
A$1,699cheapest found · Scorptec
16 GBVRAM (15.5 GB usable)
896 GB/smemory bandwidth
177 tok/smeasured, Llama 2 7B Q4_0
View at Scorptec → Price checked 20 Oct 2026, 7:40 am AEST.
Benchmark
176.85 tok/s on Llama 2 7B Q4_0, tg128 (llama.cpp llama-bench), CUDA backend. Source: ggml-org/llama.cpp/discussions/15013 scoreboard (checked 2026-10-19).
Specs: NVIDIA published specification (manually entered)
AU prices found
| Retailer | Listing | Price | Stock | Checked |
|---|---|---|---|---|
Mwave | A$1,649 | not in stock online | 20 Oct 2026, 7:38 am AEST | |
Mwave | A$1,664 | not in stock online | 20 Oct 2026, 7:38 am AEST | |
Mwave | A$1,679 | not in stock online | 20 Oct 2026, 7:38 am AEST | |
Mwave | A$1,699 | not in stock online | 20 Oct 2026, 7:38 am AEST | |
Scorptec | A$1,699 | in stock | 20 Oct 2026, 7:40 am AEST | |
Mwave | A$1,737 | not in stock online | 20 Oct 2026, 7:38 am AEST | |
Scorptec | A$1,749 | in stock | 20 Oct 2026, 7:40 am AEST | |
Scorptec | A$1,749 | in stock | 20 Oct 2026, 7:40 am AEST | |
Scorptec | A$1,749 | not in stock online | 20 Oct 2026, 7:40 am AEST | |
Mwave | A$1,789 | in stock | 20 Oct 2026, 7:38 am AEST | |
Mwave | A$1,799 | not in stock online | 20 Oct 2026, 7:38 am AEST | |
Mwave | A$1,799 | in stock | 20 Oct 2026, 7:38 am AEST |
Community measured
No reader results yet. Be the first to submit one →
Which models fit? (8k context)
| Model | Quant | Needs / usable | Fit | Est. tok/s |
|---|---|---|---|---|
Q4_K_M | 464.4 / 15.5 GB | no | – | |
Q4_K_M | 499.4 / 15.5 GB | no | – | |
Q4_K_M | 20.0 / 15.5 GB | no | – | |
Q4_K_M | 348.5 / 15.5 GB | no | – | |
Q4_K_M | 15.7 / 15.5 GB | no | – | |
Q4_K_M | 45.9 / 15.5 GB | no | – | |
Q4_K_M | 22.5 / 15.5 GB | no | – | |
Q4_K_M | 305.3 / 15.5 GB | no | – | |
MiniCPM V 4.7 1B new | Q4_K_M | 2.0 / 15.5 GB | fits | 714 estimate |
Q4_K_M | 464.4 / 15.5 GB | no | – | |
Q4_K_M | 195.9 / 15.5 GB | no | – | |
GLM 5.3 (MoE) new | Q4_K_M | 458.4 / 15.5 GB | no | – |
Q4_K_M | 110.3 / 15.5 GB | no | – | |
Q4_K_M | 18.4 / 15.5 GB | no | – | |
Q4_K_M | 20.2 / 15.5 GB | no | – | |
Q4_K_M | 9.1 / 15.5 GB | fits | 92 estimate | |
Q4_K_M | 23.0 / 15.5 GB | no | – | |
Q4_K_M | 81.4 / 15.5 GB | no | – | |
Q4_K_M | 22.1 / 15.5 GB | no | – | |
Q4_K_M | 17.2 / 15.5 GB | no | – | |
Q4_K_M | 6.0 / 15.5 GB | fits | 138 estimate | |
Q4_K_M | 7.1 / 15.5 GB | fits | 115 estimate | |
Q4_K_M | 4.1 / 15.5 GB | fits | 237 estimate | |
Q4_K_M | 77.0 / 15.5 GB | no | – | |
Q4_K_M | 20.4 / 15.5 GB | no | – | |
gpt-oss 20B (MoE) older | MXFP4 | 12.4 / 15.5 GB | fits | 351 estimate |
gpt-oss 120B (MoE) older | MXFP4 | 63.5 / 15.5 GB | no | – |
Qwen 3 4B older | Q4_K_M | 4.6 / 15.5 GB | fits | 275 estimate |
Qwen 3 8B older | Q4_K_M | 7.2 / 15.5 GB | fits | 135 estimate |
Qwen 3 32B older | Q4_K_M | 23.0 / 15.5 GB | no | – |
Llama 3.1 8B Instruct older | Q4_K_M | 6.9 / 15.5 GB | fits | 138 estimate |
Mistral 7B Instruct v0.3 older | Q4_K_M | 6.5 / 15.5 GB | fits | 153 estimate |