Ant Group, China · Released October 13, 2026New
Ling-3.1-flash has 572.1B parameters (24B active per token). In BF16 its weights alone take 1050 GiB; quantized to 4-bit, about 317 GiB. Each 32K-token request adds 0.32 GiB of KV cache. Cheapest way to run it today: 7 × RTX A6000 (48 GB) at $2.25 per hour. It calls tools and has a thinking mode you can switch off.
Size your own setup in the calculator →Weights plus the KV cache of one request (BF16 cache) and the model's fixed recurrent state, in GiB. Add 1 to 3 GiB for the inference engine itself.
| Weights format | Weights only | + 8K context | + 32K context | + 128K context | + 256K context |
|---|---|---|---|---|---|
| FP8 (native) | 527 | 527 | 527 | 528 | 530 |
| BF16 / FP16 | 1050 | 1051 | 1051 | 1052 | 1053 |
| FP8 / INT8 | 526 | 527 | 527 | 528 | 529 |
| 4-bit (GGUF Q4_K_M) | 317 | 318 | 318 | 319 | 320 |
Exact size of each quantization in AtomicChat/Ling-3.1-flash-GGUF (opens in a new tab), and the smallest GPU that runs it with llama.cpp, Ollama or LM Studio and an 8K-token context.
| Quantization | File size | Smallest GPU |
|---|---|---|
| Q8_0 | 560 GiB | More than 96 GB: see the calculator |
| Q6_K | 445 GiB | More than 96 GB: see the calculator |
| Q5_K_M | 373 GiB | More than 96 GB: see the calculator |
| Q4_K_M | 330 GiB | More than 96 GB: see the calculator |
| IQ4_XS | 315 GiB | More than 96 GB: see the calculator |
| IQ1_S | 132 GiB | More than 96 GB: see the calculator |
| Quantization | File size | Smallest GPU |
|---|---|---|
| IQ4_XXS | 290 GiB | More than 96 GB: see the calculator |
| IQ3_M | 258 GiB | More than 96 GB: see the calculator |
| IQ3_XXS | 239 GiB | More than 96 GB: see the calculator |
| IQ2_M | 203 GiB | More than 96 GB: see the calculator |
| IQ2_XS | 185 GiB | More than 96 GB: see the calculator |
| IQ2_XXS | 162 GiB | More than 96 GB: see the calculator |
| IQ1_M | 150 GiB | More than 96 GB: see the calculator |
Read every day from Hugging Face. Split files are added together; vision projectors are left out.
Fewest GPUs that fit the weights plus one 8K-token request. Data-center GPUs use vLLM (90% of memory usable), the others llama.cpp. Several GPUs split the model with tensor or pipeline parallelism.
| GPU | FP8 (native) | BF16 / FP16 | 8-bit | 4-bit |
|---|---|---|---|---|
| GeForce RTX 4090 24 GB | Not on this GPU | 64 (several machines) | 32 (several machines) | 16 (several machines) |
| GeForce RTX 5090 32 GB | Not on this GPU | 48 (several machines) | 24 (several machines) | 16 (several machines) |
| RTX PRO 6000 Blackwell 96 GB | Not on this GPU | 16 (several machines) | 6 | 4 |
| Mac M5 Max 128 GB 96 GB usable | Not on this GPU | 16 (several machines) | 6 | 4 |
| Ryzen AI Max+ 395 128 GB 96 GB usable | Not on this GPU | 16 (several machines) | 6 | 4 |
| DGX Spark (GB10) 128 GB | Not on this GPU | 16 (several machines) | 5 | 3 |
| NVIDIA H100 80 GB | 8 | 16 (several servers) | 8 | 5 |
| NVIDIA H200 141 GB | 5 | 16 (several servers) | 5 | 3 |
| NVIDIA B200 (HGX) 180 GB | 4 | 7 | 4 | 3 |
8-bit: FP8 with vLLM (INT8 on GPUs without FP8), Q8_0 with llama.cpp. 4-bit: AWQ / GPTQ with vLLM, Q4_K_M with llama.cpp, both counted at the Q4_K_M size. Not on this GPU: the official FP8 weights do not run there (llama.cpp needs a GGUF version, and vLLM needs a GPU with FP8 support).
For each GPU: the fewest cards that fit Ling-3.1-flash with one 8K-token request, the most faithful weight format at that count, and the cheapest on-demand price (2026-10-16).
| Setup | Per hour | Weights | Provider | Per month | Rent |
|---|---|---|---|---|---|
| 7 × RTX A6000 48 GB | $2.25 | Q4_K_M | Vast.ai | $1,640.31 | Rent 7 × RTX A6000 on Vast.ai (opens in a new tab) → |
| 7 × RTX 6000 Ada 48 GB | $3.94 | Q4_K_M | Vast.ai | $2,876.93 | Rent 7 × RTX 6000 Ada on Vast.ai (opens in a new tab) → |
| 5 × NVIDIA A100 80GB | $4.68 | 4-bit | Vast.ai | $3,412.75 | Rent 5 × NVIDIA A100 80GB on Vast.ai (opens in a new tab) → |
| 2 × AMD MI300X 192 GB | $4.78 | 4-bit | RunPod | $3,489.40 | Rent 2 × AMD MI300X on RunPod (opens in a new tab) → |
| 4 × RTX PRO 6000 Blackwell 96 GB | $5.34 | Q4_K_M | Vast.ai | $3,895.28 | Rent 4 × RTX PRO 6000 Blackwell on Vast.ai (opens in a new tab) → |
| 3 × NVIDIA H200 141 GB | $8.69 | 4-bit | Vast.ai | $6,340.05 | Rent 3 × NVIDIA H200 on Vast.ai (opens in a new tab) → |
| 2 × AMD MI355X / MI350X 288 GB | $10.98 | 4-bit | RunPod | $8,015.40 | Rent 2 × AMD MI355X / MI350X on RunPod (opens in a new tab) → |
| 4 × NVIDIA H100 NVL 94 GB | $11.07 | 4-bit | Vast.ai | $8,082.56 | Rent 4 × NVIDIA H100 NVL on Vast.ai (opens in a new tab) → |
On-demand prices from AWS, Google Cloud, Azure, RunPod, Vast.ai, Oracle Cloud, OVHcloud, Scaleway, Verda, Akash and UpCloud, checked every hour. Serving many users needs more KV cache, so more memory: size it in the calculator. Some links are affiliate links: StudioTV may earn a commission, at no extra cost to you.
AWS, Google Cloud and UpCloud prices: Spare Cores (opens in a new tab), CC BY-SA 4.0 (opens in a new tab), shared under the same license.
No GPU to rent: these providers serve Ling-3.1-flash and bill per token. Speeds are measured by Hugging Face for one request.
| Provider | Output, per 1M tokens | Input, per 1M | Speed | First token |
|---|---|---|---|---|
| Novita | $0.900 | $0.300 | 78 tokens/s | 1.1 s |
Prices and speeds from Hugging Face Inference Providers (opens in a new tab), read every day (October 16, 2026). An API wins at low volume; renting GPUs wins when they stay busy: the calculator gives the break-even.
Reasoning. Thinking mode you can switch on or off. With vLLM or SGLang, set it per request: "chat_template_kwargs": {"enable_thinking": false} turns it off, true on. Reasoning tokens take context and KV cache like any other token: leave room for them.
Architecture: 72 layers · MLA (512 + 64 latent) on 9 · 63 Kimi Delta Attention. Its KV cache is 8.0× smaller than if every layer used full attention, because only 9 of its 72 layers keep a KV cache; the others hold a small fixed-size state. Only 24B parameters are active per token, but all 572B must sit in memory: VRAM depends on the total, speed on the active part.
| Context per request | BF16 cache | FP8 cache |
|---|---|---|
| 8K tokens | 0.08 GiB | 0.04 GiB |
| 32K tokens | 0.32 GiB | 0.16 GiB |
| 128K tokens | 1.27 GiB | 0.63 GiB |
| 256K tokens | 2.53 GiB | 1.27 GiB |
Plus 261 MiB of recurrent state per request, whatever the context.
In BF16 the weights alone take 1050 GiB (1128 GB). In FP8 that is 526 GiB, and about 317 GiB with 4-bit quantization (Q4_K_M); the official FP8 checkpoint is 527 GiB. Each request then adds KV cache: 0.08 GiB at 8K tokens and 0.32 GiB at 32K (BF16 cache).
No. Even in 4-bit it needs 16 RTX 4090s, spread over several machines. In 4-bit it needs 5 H100 GPUs.
No. Even in 4-bit it needs 16 RTX 5090s, spread over several machines.
Yes. In 4-bit (Q4_K_M with llama.cpp) it needs a Mac with 512 GB of unified memory, of which macOS gives the GPU about 384 GB, for one request with an 8K-token context. In 8-bit it does not fit on a single Mac.
No. Even in 4-bit it needs 3 DGX Spark units or 4 Ryzen AI Max+ 395 machines (128 GB), linked over the network.
On 2026-10-16, the cheapest on-demand setup is 7 × RTX A6000 (48 GB) with Q4_K_M weights on Vast.ai, at $2.25 per hour (about $1,640.31 per month). Next: 7 × RTX 6000 Ada (48 GB) with Q4_K_M weights on Vast.ai, at $3.94 per hour. Sized for one 8K-token request; prices are checked every hour.
With one request and an 8K-token context: 16 in BF16, 8 in FP8 and 5 in 4-bit (vLLM, 90% of memory usable). Above 8 GPUs, that means several servers. Serving many users at once needs more memory for their KV caches: the calculator sizes that for you.
Yes, Ling-3.1-flash supports tool (function) calling: it can ask your application to run a function and use the result.
Thinking mode you can switch on or off. With vLLM or SGLang, set it per request with "chat_template_kwargs": {"enable_thinking": false} (or true). Reasoning tokens count toward the context and the KV cache.
For a model card or a README: the 4-bit size and the smallest GPU it fits on, linked to this page. Data as JSON: /api/models/ling-3-1-flash.json.
Ant Group, China · Released October 13, 2026[](https://studiotvai.com/vram-requirements/ling-3-1-flash)Same family (large moe): MiniMax M2.7 · DeepSeek R1 · Kimi K2.6 · GLM-5.3 · GLM-5.3-Flash · DeepSeek V4 Flash · DeepSeek V4 Pro · DeepSeek V4.1 Flash · MiMo-V2.6 Pro · MiMo-V2.6 Flash · Kimi K3 · Mistral Large 4 · DeepSeek V4 Flash Vision Exp · Qwen3.8 2.4T-A95B · Ornith 1.5 397B · Atria Dawn Preview · Intern S2 397B · DeepSeek V3 · DeepSeek V3.1 · Kimi K2.7 Code · Kimi K2.5 · Kimi K2 · GLM-5.2 · GLM-5.1 · GLM-5 · MiniMax M2.5 · MiniMax M2 · MiMo-V2.5 · Ornith-1.0 397B · Qwen3 235B-A22B · Qwen3.5 397B-A17B · GLM-4.7 · GLM-4.5 · Step-5-Preview
Llama 3.1 8B · Mistral Small 3.2 24B · Qwen3 32B · Llama 3.3 70B · Llama 3.1 405B · Qwen3.5 4B · Qwen3.5 9B · Qwen3.8 27B · Qwen3.6 35B-A3B · Qwen3.5 122B-A10B · Qwen3.8-Flash-Next · LFM2.5 8B-A1B · LFM2 24B-A2B · Gemma 4 12B · Gemma 4 31B · Gemma 4 26B-A4B · gpt-oss 20B · gpt-oss 120B · Llama 4 Scout 17B-16E · Llama 4 Maverick 17B-128E · Kolibri-1 · Nemotron 3 Nano 30B-A3B · Nemotron 3 Super 120B-A12B · MiniMax M2.7 · DeepSeek R1 · Kimi K2.6 · GLM-5.3 · GLM-5.3-Flash · DeepSeek V4 Flash · DeepSeek V4 Pro · DeepSeek V4.1 Flash · MiMo-V2.6 Pro · MiMo-V2.6 Flash · Kimi K3 · Mistral Large 4 · Qwen3-Coder-Next 80B-A3B · Qwen3-Coder 30B-A3B · Qwen3-Coder 480B-A35B · DeepSeek V4 Flash Vision Exp · Qwen3.8 2.4T-A95B · Ornith 1.5 397B · Ornith 1.5 35B-A3B · Ornith 1.5 9B · Atria Dawn Preview · Intern S2 397B · Clef · Clef Flash · D1 3B · Humanizer · JEV 27B VL · Mellum2.1 12B-A2.5B · LightOnOCR-3 4B · Spark-X2.5 4B · Agnes-3.0-Qwen · Youtu-Parsing-Omni · Agnes-3.0-Flash · Mistral 7B v0.3 · Mistral Small 3.1 24B · Mistral Nemo · Ministral 8B · DeepSeek V3 · DeepSeek R1 Qwen3 8B · DeepSeek R1 Distill Qwen 32B · DeepSeek V3.1 · DeepSeek R1 Distill Qwen 14B · DeepSeek R1 Distill Qwen 7B · DeepSeek R1 Distill Llama 8B · Qwen3 4B · Qwen3 8B · Qwen2.5 7B · Qwen3 14B · Qwen2.5 3B · Qwen3.6 27B · Qwen3.5 27B · Qwen2.5-Coder 7B · Qwen2.5 32B · Qwen2.5-Coder 14B · Qwen3.5 35B-A3B · Qwen3 30B-A3B · Qwen2.5 14B · Qwen2.5-Coder 32B · Kimi K2.7 Code · Kimi K2.5 · Kimi K2 · GLM-5.2 · GLM-5.1 · GLM-5 · MiniMax M2.5 · MiniMax M2 · MiMo-V2.5 · Ornith-1.0 397B · Qwen2.5-Coder 3B · Qwen3 235B-A22B · Qwen3.5 397B-A17B · Gemma 4 E4B · Gemma 4 E2B · Nemotron 3 Nano 4B · Nemotron 3.5 Lightning 30B-A3B · MiniCPM-V-4.7 35B-A3B · LFM2.5 24B-A2B · GLM-4.7 · Qwen3-VL 8B · Qwen2.5-VL 7B · Qwen3-VL 4B · Qwen2.5-VL 3B · Qwen2.5-VL 32B · Qwen2-VL 7B · GLM-4.5-Air · GLM-4.5 · phi-4 · Phi-3.5-mini · Phi-4-mini · Phi-3-mini-128k · Unlimited-OCR · Step-5-Preview
Estimates, not guarantees: computed from the official config.json with the same engine as the LLM VRAM Calculator. Real usage depends on your engine version and settings.