Ling-3.1-flash on local hardware
Estimated comfort for repository coding with Aider on a MacBook Pro M4 Pro 48GB.
Ling-3.1-flashOpen source
inclusionAI · mit · 2,468 downloads
Mixture of expertsReasoning / thinking mode
- Parameters
- 572B exact
- Active per token
- ≈32B (8 of 512 experts)
- Context window
- 256K tokens
- Attention
- Multi-head latent attention (compressed KV)
- Layers · KV heads · head dim
- 72 · 64 · 128
- KV cache
- 81 KB/token · 2.5 GB at 32K
- Tool calling
- good (from chat template)
- Thinking
- good (from chat template)
- Vision
- basic (inferred — no supporting signal)
- Commercial use
- allowedRetain the required license and attribution notices
- Import confidence
- medium · capabilities estimated
Weights by quantization
Q3_K_M
256 GB
gguf file size
Q4_K_M
333 GB
gguf file size
Q5_K_M
406 GB
gguf file size
Q6_K
467 GB
gguf file size
Q8_0
567 GB
gguf file size
- • vision: basic, inferred from missing signals (low confidence).
- • Capability ratings are estimated from model size and name — this model hasn't been assessed by hand.
- • GGUF sizes come from community conversions matched by model name; equivalence and runtime support are not verified. MLX conversion availability is unverified.
Ling-3.1-flash — 3-bit
Does not fit in memory
Needs ≈239 GB but only ≈37 GB is available after the OS, other apps and Aider.
High confidence