// Ant Ling / Ring / Ming · Text LLM
Ling-3.0-tiny: VRAM & local deployment
7.9B total · 1.3B active (MoE) parameters. About 6.0 GB at 4-bit and 18.1 GB at BF16 with one 8K-token request.
released 2026-08-10licence: Commercial OKmemory figures are estimates
01Specs
| Publisher | Ant Group inclusionAI · inclusionAI |
|---|---|
| Family / series | Ant Ling / Ring / Ming · Ling-3.0-tiny |
| Type | Text LLM (text-generation) |
| Parameters | 7.9B total · 1.3B active (MoE) Source: official model card“7.9B total parameters and only 1.3B activated parameters per token” |
| Context length | 128K tokens (131,072)max_position_embeddings in config.json |
| Architecture | 24 layers · MLA · MoE |
| Released weights | BF16 · 15.8 GBdownload |
| Licence | MIT · Commercial OKPermissive licence (Apache-2.0 / MIT / CC-BY / OpenMDW): commercial use allowed. |
| Released | 2026-08-10 (Hugging Face repo created) |
02How much VRAM it needs (estimate)
Total = weights + 10% + 0.5GB overhead + KV cache for one request (FP16 KV). More concurrent users multiply the KV column.
| Precision | Weights | Total @ 8K | Total @ 32K | Total @ 128K | Smallest GPU | Mac |
|---|---|---|---|---|---|---|
| BF16 / FP16 (original) | 15.8 GB | 18.1 GB | 18.8 GB | 21.5 GB | RTX 4090 24GB | Mac 36GB |
| FP8 / INT8 / Q8_0 | 8.4 GB | 10.0 GB | 10.6 GB | 13.3 GB | 12GB GPU | Mac 24GB |
| 4-bit (Q4_K_M / AWQ / GPTQ-Int4) | 4.8 GB | 6.0 GB | 6.7 GB | 9.4 GB | 8GB GPU | Mac 16GB |
KV cache per request: 8K 0.2 GB · 32K 0.9 GB · 128K 3.6 GB
Bits per weight: BF16 16, FP8 8.5, 4-bit 4.85. “Fits” = total ≤ 90% of memory; Mac GPU can use ~2/3 of unified memory (≤36GB) or ~3/4 (larger). For MoE, all experts must be in memory, so size by total parameters. Method →
03How to run it
Only runtimes the publisher or the runtime’s official library names for this model.
- vLLM — mentioned in the official model card
- SGLang — mentioned in the official model card
- Ollama — mentioned in the official model card
- MLX — mentioned in the official model card
- llama.cpp — official GGUF weights
04Official repositories & quantised builds
5 official repo(s) from inclusionAI.
- Ling-3.0-tiny-GGUF GGUF
- Ling-3.0-tiny main
- Ling-3.0-tiny-int4 INT4
- Ling-3.0-tiny-fp8 FP8
- Ling-3.0-tiny-base Base
05Other Ant Ling / Ring / Ming models
All 29 →| Model | Params | 4-bit total | Released |
|---|---|---|---|
| Ling-3.0-tiny-UBTree | 183M | 1.0 GB | 2026-10-21 |
| Ling-3.0-flash-UBTree | 482M | 1.5 GB | 2026-10-14 |
| Ming-3.0-asr-tiny | 8.87B | 6.4 GB | 2026-10-14 |
| Ling-3.1-flash | 560B / 25B | 375 GB | 2026-10-13 |
| Ming-3.0-tts-0.5B | 1.48B | 1.5 GB | 2026-09-22 |
| Ming-3.0-tts-tiny | 8.98B / 1.3B | 6.5 GB | 2026-09-22 |
| Ming-Image-0.1-Design | 6.16B | 4.6 GB | 2026-09-17 |
| Ling-3.0-flash-VL | 124B / 5.5B | 83.6 GB | 2026-09-04 |
06Sources
Data checked 2026-09-24. Parameters, context and licence are copied from the official repository; memory figures are our estimates.