GPU VRAM catalog
Compare GPU memory tiers for local AI
Start with the amount of VRAM your workload needs, then compare platform support, power, and exact-card requirements inside that tier.
Start with the constraint
Choose a VRAM tier before a product name
Product tier does not override physical memory capacity. Use the calculator to establish a range, then jump to the closest tier below.
How to narrow the catalog
Estimate memory
Include model weights, runtime overhead, context or image settings, and safety margin.
Check the platform
Confirm CUDA, ROCm, Intel, operating-system, and framework support for the workflow.
Compare within a tier
Use power, memory configuration, architecture, and source confidence to refine the shortlist.
Verify the exact card
Partner-card connectors, cooling, dimensions, and factory settings can differ.
8GB profiles
8GB entry planning tier
RTX 5060
Source-backed planning profile for 8GB mainstream Blackwell creator and AI-coding research.
View planning profile →RTX 5050
Source-backed planning profile for entry-level 8GB Blackwell GPU research.
View planning profile →12GB profiles
12GB planning tier
RTX 5070
Source-backed planning profile for 12GB Blackwell local AI and creator workflow research.
View planning profile →RTX 4070 Super
Seed record for later sourced comparison and AI workflow testing.
View planning profile →RTX 4070
Source-backed planning profile for creator-focused local AI workloads.
View planning profile →RTX 3060 12GB
Source-backed planning profile for entry-level local LLM and Stable Diffusion workflows.
View planning profile →16GB profiles
16GB headroom tier
RTX 5080
Source-backed planning profile for Blackwell 16GB local AI and creator workflow research.
View planning profile →RTX 5070 Ti
Source-backed planning profile for 16GB Blackwell local AI and creator workflow research.
View planning profile →RTX 5060 Ti 16GB
Source-backed planning profile for 16GB mainstream Blackwell local AI research.
View planning profile →RTX 4080 Super
Source-backed planning profile for high-end local AI creator workflows.
View planning profile →RTX 4070 Ti Super
Source-backed planning profile for higher-memory local AI workloads.
View planning profile →RTX 4060 Ti 16GB
Source-backed planning profile for mid-range local AI and image workflow research.
View planning profile →Intel Arc A770 16GB
Source-backed planning profile for Intel-based local AI and creator workflow research.
View planning profile →24GB+ profiles
24GB and higher-memory tier
RTX 5090
Source-backed planning profile for next-generation local AI hardware research.
View planning profile →RTX 4090
Source-backed planning profile for high-memory local AI workflows.
View planning profile →RTX 3090
Source-backed planning profile for high-VRAM local LLM workstation research.
View planning profile →RX 7900 XTX
Source-backed planning profile for AMD-based local AI workstation exploration.
View planning profile →Do not choose by VRAM alone
Capacity creates the shortlist. Runtime support determines whether that capacity is practical for your software stack.