INTELLECT 3 — Hardware Requirements & GPU Compatibility
ChatReasoning
INTELLECT 3 is a 106.9B-parameter open language model from PrimeIntellect. It supports a context window of up to 131,072 tokens. At Q4_K_M it needs about 64.54 GB of VRAM — see which GPUs and Macs can run it below.
INTELLECT 3 runs on a 96 GB+ Mac or a workstation GPU
Best local pick → Mac Studio (M3 Ultra, 96GB) · Q4_K_M · 65.4 GB · ~47 tok/s
Macs from 64 GB (Q3_K_M · ~52 tok/s)
MIT
Not set up for it?
Reference GPU · RTX 4090 24 GB
Estimates, ±20% · how we estimate
Run INTELLECT 3
One recommended way. Every other method, OS and quant is in the step-by-step guide.
Recommended
vllm serve PrimeIntellect/INTELLECT-3 --max-model-len 16384vLLM repo with its own quantisation
Download: Weights on Hugging Face
Other ways
each opens step-by-stepRun in cloud
Fits on A100 80GB (14 GB headroom) · Q4_K_M
- Generation speed
- ~130 tok/s
- generation speed
- Cost per 1M output tokens
- $2.85
- per 1M output tokens
How Much VRAM Does INTELLECT 3 Need?
Select a quantization level; the listed hardware adheres to it.
| Quantization | VRAM | +32K | On RTX 4090 (reference) | |
|---|---|---|---|---|
| 45.8 GB | 47.8 GB | 2-bit · K-quant — 2-bit quantization with K-quant improvements | Does not fit: needs 46.7 GB | |
| 52.5 GB | 54.5 GB | 3-bit medium — 3-bit medium quantization | Does not fit: needs 53.4 GB | |
| 64.5 GB | 66.5 GB | 4-bit medium — 4-bit medium quantization — most popular sweet spot | Does not fit: needs 65.4 GB | |
| 76.6 GB | 78.5 GB | 5-bit medium — 5-bit medium quantization — good quality/size tradeoff | Does not fit: needs 77.5 GB | |
| 88.6 GB | 90.5 GB | 6-bit, very good — 6-bit quantization, very good quality | Does not fit: needs 89.5 GB | |
| 107.3 GB | 109.2 GB | 8-bit, near-lossless — 8-bit quantization, near-lossless | Does not fit: needs 108.2 GB | |
| 214.1 GB | 216.1 GB | 16-bit · bfloat — Brain floating point 16 — preferred for training | Does not fit: needs 215 GB |
Selected Q4_K_M
4-bit medium quantization65.4GBat 16K context
- Weights
- 64.1 GB
- KV cache
- 1.0 GB
- Overhead
- 0.3 GB
Doesn't fit RTX 4090 (reference) · needs 41.4 GB more
Which Hardware Can Run INTELLECT 3?
Q4_K_M · 64.5 GBINTELLECT 3 (Q4_K_M) requires 64.5 GB of VRAM to load the model weights. For comfortable inference with headroom for KV cache and system overhead, 84+ GB is recommended. Using the full 131K context window can add up to 8.1 GB, bringing total usage to 72.6 GB. No consumer GPU has enough memory. 30 devices with unified memory can run INTELLECT 3, including Mac Studio (M3 Ultra, 256GB), Mac Studio (M3 Ultra, 512GB), Mac Studio (M3 Ultra, 96GB).
64–80 GB
Q4_K_M · best fit Q3_K_M (53.4 GB)
3 Mac classes · 1 device family
M2 Ultra*64 GB~52Simulate output on Mac Pro M2 Ultra (64 GB)M5 Max64 GB~45Simulate output on Mac Studio M5 Max (64 GB)Orin~23Simulate output on NVIDIA Jetson AGX Orin 64GB96 GB
fits · 30.6 GB free · up to 128K ctx
2 Mac classes
M3 Ultra*96 GB~47Simulate output on Mac Studio (M3 Ultra, 96GB)M3 Max*96 GB~24Simulate output on MacBook Pro 14" M3 Max (96 GB)
By VRAM tier · typical GPUs and Macs at Q4_K_M
How good is INTELLECT 3?
llmrun Score blends public benchmark boards into one number; 100 = the average open model.
Not enough data yet for an llmrun Score.
A Score needs results on enough independent boards. How the Composite Score is built
Benchmarks
This is the single board showing results for this model; badges show who ran the test.
- LMArena Text
LMArena · Oct 2026Arena
Arena1355.8
#72/187 open
Similar models
5| Model | Score | VRAM | tok/s | Why it is similar |
|---|---|---|---|---|
| Llama 4 Scout 17B 16E InstructMeta | 81.8 | 77 GB | —needs > 24 GB | same size |
| C4ai Command A 03 2025Cohere | 79.8 | 78.7 GB | —needs > 24 GB | same size |
| GLM 4.6VZ.ai | — | 66 GB | —needs > 24 GB | same size |
| C4ai Command R Plus 08 2024Cohere | — | 73.8 GB | —needs > 24 GB | same size |
| Qwen1.5 110BAlibaba | — | 72.4 GB | —needs > 24 GB | same size |
Fine-tunes & variants(1)
Specifications
- Publisher
- PrimeIntellect
- Parameters
- 106.9B
- Active parameters
- 11.3B per token
- Architecture
- Glm4MoeForCausalLM
- Context Length
- 131,072 tokens
- Vocabulary Size
- 151,552
- Release Date
- 2025-11-26
- License
- MIT
Frequently Asked Questions
How much VRAM does INTELLECT 3 need?
INTELLECT 3 requires 64.5 GB of VRAM at Q4_K_M, or 214.1 GB at BF16. Full 131K context adds up to 8.1 GB (72.6 GB total).
VRAM = Weights + KV Cache + Overhead
Weights = 106.9B × 4.8 bits ÷ 8 = 64.1 GB
KV Cache + Overhead ≈ 0.4 GB (at 2K context + ~0.3 GB framework)
Fit ratings and hardware model lists check this model with room for a 16K-token context, which needs a little more memory.
KV Cache + Overhead ≈ 8.5 GB (at full 131K context)
VRAM usage by quantization
Can NVIDIA GeForce RTX 5090 run INTELLECT 3?
No — INTELLECT 3 requires at least 45.8 GB at Q2_K, which exceeds the NVIDIA GeForce RTX 5090's 32 GB of VRAM.
What's the best quantization for INTELLECT 3?
For INTELLECT 3, Q4_K_M (64.5 GB) offers the best balance of quality and VRAM usage. Q5_K_M (76.6 GB) provides better quality if you have the VRAM. The smallest option is Q2_K at 45.8 GB.
VRAM requirement by quantization
★ Recommended — best balance of quality and VRAM usage.
Can I run INTELLECT 3 on a Mac?
Yes — Mac Pro M2 Ultra (64 GB) and 41 other Macs can run INTELLECT 3, the smallest at Q3_K_M. Apple Silicon uses unified memory, so the model shares RAM with the system. At Q4_K_M you need at least 64.5 GB of usable unified memory (RAM minus macOS overhead), from Mac Studio (M3 Ultra, 96GB).
Can I run INTELLECT 3 locally?
Yes — INTELLECT 3 can run locally on consumer hardware. Popular tools include Ollama, LM Studio, and llama.cpp.
How fast is INTELLECT 3?
At Q4_K_M, INTELLECT 3 can reach ~94 tok/s on AMD Instinct MI350X. Speed depends mainly on GPU memory bandwidth. Real-world results typically within ±20%.
tok/s = 1000 ÷ (active GB ÷ (bandwidth GB/s × efficiency) × 1000 + layers × routing ms)
Mixture-of-Experts: just the active experts get read for each token, along with a fixed routing cost per layer.
Example: NVIDIA B200 → 6.9 GB active ÷ (8000 × 0.65) = 1.32 ms, plus 46 layers × 0.055 ms = 2.53 ms, so 1000 ÷ 3.85 ms = ~260 tok/s
Estimated speed at Q4_K_M (64.5 GB)
Real-world results typically within ±20%. Speed depends on batch size, quantization kernel, and software stack.
What's the download size of INTELLECT 3?
At Q4_K_M, the download is about 64.11 GB. The full-precision BF16 version is 213.70 GB. The smallest option (Q2_K) is 45.41 GB.
Which GPUs can run INTELLECT 3?
No single consumer GPU has enough VRAM to run INTELLECT 3 at Q4_K_M (64.5 GB). Multi-GPU or professional hardware is required.
Which devices can run INTELLECT 3?
30 devices with unified memory can run INTELLECT 3 at Q4_K_M (64.5 GB), including ASUS Ascent GX10, Beelink GTR9 Pro (Ryzen AI Max+ 395, 128 GB), Framework Desktop (Ryzen AI Max+ 395, 128 GB), GMKtec EVO-X2 (Ryzen AI Max+ 395, 128 GB). Apple Silicon Macs use unified memory shared between CPU and GPU, making them well-suited for local LLM inference.