Arc A770 16GB

Best local AI models for this GPU

VRAM
16 GB
Bandwidth
560 GB/s
Vendor
Intel
Compatible Models
16

How it compares

Arc A770 16GB vs GPUs at nearby launch prices. Estimates from memory bandwidth, not benchmarks.

Est. speed, Qwen3 8B (Q4_K_M)

RTX 2060~48 tok/s · 6 GBRX 5700~65 tok/s · 8 GBArc A770 16GB~81 tok/s · 16 GBRX 6600 XT~37 tok/s · 8 GBGTX 1070~37 tok/s · 8 GB

Models at 10+ tok/s (est.)

RTX 2060~5 models · $349RX 5700~10 models · $349Arc A770 16GB~16 models · $349RX 6600 XT~10 models · $379GTX 1070~10 models · $379

Top Compatible Models

Top 10 models ranked by performance. Click any model to expand details or view full compatibility.

Memory Usage10.6 GB
Weights
KV Cache
Overhead
Context Length131,072 tokens
Install with Ollama:
ollama run qwen3:14b
View full compatibility
Memory Usage10.8 GB
Weights
KV Cache
Overhead
Context Length131,072 tokens
Install with Ollama:
ollama run deepseek-r1:14b
View full compatibility
Memory Usage10.6 GB
Weights
KV Cache
Overhead
Context Length16,384 tokens
Install with Ollama:
ollama run phi4
View full compatibility
Memory Usage13.2 GB
Weights
KV Cache
Overhead
Context Length16,384 tokens
Install with Ollama:
ollama run gemma3:12b
View full compatibility
Memory Usage10.8 GB
Weights
KV Cache
Overhead
Context Length131,072 tokens
Install with Ollama:
ollama run qwen2.5:14b
View full compatibility
Memory Usage9.2 GB
Weights
KV Cache
Overhead
Context Length131,072 tokens
Install with Ollama:
ollama run mistral-nemo
View full compatibility
Memory Usage7.4 GB
Weights
KV Cache
Overhead
Context Length131,072 tokens
Install with Ollama:
ollama run qwen3:8b-q5_k_m
View full compatibility
Memory Usage7.4 GB
Weights
KV Cache
Overhead
Context Length131,072 tokens
Install with Ollama:
ollama run deepseek-r1:8b-q5_k_m
View full compatibility
Memory Usage7.3 GB
Weights
KV Cache
Overhead
Context Length131,072 tokens
Install with Ollama:
ollama run llama3.1:8b-q5_k_m
View full compatibility
Memory Usage7.2 GB
Weights
KV Cache
Overhead
Context Length131,072 tokens
Install with Ollama:
ollama run qwen2.5:7b-q5_k_m
View full compatibility