IQuest Q1 — Hardware Requirements & GPU Compatibility
ChatFunctionsIQuest-Q1 is IQuest's Mixture-of-Experts language model featuring 320 billion total parameters, with roughly 15 billion active ones, designed for agentic coding, reasoning, and multi-step tool use. The card details its structure: 88 layers, 256 experts with 8 active per token, a hybrid attention pattern combining three sliding-window layers to one full-attention layer, plus multi-token prediction layers. It handles text only in English and Chinese. Even when quantized, the complete weights require a multi-GPU server or a very large unified-memory machine, although only the active experts calculate each token. The model supports a context window of 524,288 tokens. Released under the custom IQuest-Q1 license instead of a standard open-source one, you must check the repository's LICENSE file before commercial use. It came out in September 2026. The card compares it to DeepSeek-V4-Flash and DeepSeek-V4-Pro on agentic and coding benchmarks.
IQuest Q1 runs on a 256 GB+ Mac or a datacenter GPU
Best local pick → Mac Studio M5 Ultra (256 GB) · Q4_K_M · 195.4 GB · ~36 tok/s
Macs from 192 GB (Q3_K_M · ~33 tok/s)
Other
Not set up for it?
Reference GPU · RTX 4090 24 GB
Estimates, ±20% · how we estimate
Run IQuest Q1
One recommended way. Every other method, OS and quant is in the step-by-step guide.Each way opens a step-by-step guide for your OS and quant.
Recommended
ollama run hf.co/xhxlb/IQuest-Q1-GGUF:Q4_K_MQ4_K_M runs on one Instinct MI325X 256 GB · 16K context
Download: Q4_K_M/IQuest-Q1-Q4_K_M-00001-of-00016.gguf194.00 GB + 15 partsxhxlb
Ways to runOther ways
each opens step-by-stepRun in cloud
Runs on B300 (288 GB) (92 GB headroom) · Q4_K_M
- Generation speed
- ~152 tok/s
- generation speed
- Cost per 1M output tokens
- $12.66
- per 1M output tokens
How Much VRAM Does IQuest Q1 Need?
Select a quantization level; the listed hardware adheres to it.
| Quantization | VRAM | +32K | Runs on | |
|---|---|---|---|---|
| 136.8 GB | 142.3 GB | 2-bit · K-quant — 2-bit quantization with K-quant improvements | Runs on H200 NVL 141 GB | |
| 156.8 GB | 162.4 GB | 3-bit medium — 3-bit medium quantization | Runs on B200 192 GB | |
| 192.9 GB | 198.4 GB | 4-bit medium — 4-bit medium quantization — most popular sweet spot | Runs on Instinct MI325X 256 GB | |
| 228.9 GB | 234.4 GB | 5-bit medium — 5-bit medium quantization — good quality/size tradeoff | Runs on Instinct MI325X 256 GB | |
| 264.9 GB | 270.5 GB | 6-bit, very good — 6-bit quantization, very good quality | Runs on Instinct MI350X 288 GB | |
| 321.0 GB | 326.5 GB | 8-bit, near-lossless — 8-bit quantization, near-lossless | Runs on 5× H100 SXM 80 GB | |
| 641.3 GB | 646.9 GB | 16-bit · bfloat — Brain floating point 16 — preferred for training | Runs on 6× H200 SXM 141 GB |
Selected Q4_K_M
4-bit medium quantization195.4GBat 16K context
- Weights
- 192.2 GB
- KV cache
- 3.0 GB
- Overhead
- 0.3 GB
Server-class: ~195.4 GB at 16K context · one Instinct MI325X 256 GB
Or .
Which Hardware Can Run IQuest Q1?
Q4_K_M · 192.9 GBIQuest Q1 (Q4_K_M) requires 192.9 GB of VRAM to load the model weights. For comfortable inference with headroom for KV cache and system overhead, 251+ GB is recommended. Using the full 524K context window can add up to 94.1 GB, bringing total usage to 287.0 GB. No consumer GPU has enough memory. 4 devices with unified memory can run IQuest Q1, including Mac Studio M5 Ultra (512 GB), Mac Studio (M3 Ultra, 512GB), Mac Studio M5 Ultra (256 GB).
192+ GB
Q4_K_M · best fit Q3_K_M (159.4 GB)
6 GPUs · 1 Mac class · 9 device families
B200~160Simulate output on B200B300~160Simulate output on B300Instinct MI350X~52Simulate output on Instinct MI350XM5 Ultra256–512 GB~38Simulate output on Mac Studio M5 Ultra (256 GB)8x H200 SXM5~166Simulate output on NVIDIA DGX H20072x B200 + 36x Grace~160Simulate output on NVIDIA GB200 NVL72
By VRAM tier · typical GPUs and Macs at Q4_K_M
Consumer GPUs (up to 32 GB) and Macs up to 128 GB: too small for this model (needs ≥ 139.4 GB)8 GB tier →12 GB tier →16 GB tier →24 GB tier →Macs →
Similar models
5| Model | Score | VRAM | tok/s | Why it is similar |
|---|---|---|---|---|
| GLM 5.3 FlashZ.ai | 124.0 | 205.2 GB | —needs > 24 GB | same size |
| MiMo V2.6 Flash MOPDXiaomi | — | 186.5 GB | —needs > 24 GB | same size |
| DeepSeek V4 Flash Vision ExpDeepSeek | — | 183.3 GB | —needs > 24 GB | same size |
| Hy3Tencent | — | 182.3 GB | —needs > 24 GB | same size |
| Step3StepFun | — | 219.1 GB | —needs > 24 GB | same size |
Specifications
- Publisher
- IQuestLab
- Parameters
- 320.3B
- Active parameters
- 14.9B per token
- Architecture
- IQuestQ1ForCausalLM
- Context Length
- 524,288 tokens
- Vocabulary Size
- 160,000
- Release Date
- 2026-09-28
- License
- Other
Frequently Asked Questions
How much VRAM does IQuest Q1 need?
IQuest Q1 requires 192.9 GB of VRAM at Q4_K_M, or 641.3 GB at BF16. Full 524K context adds up to 94.1 GB (287.0 GB total).
VRAM = Weights + KV Cache + Overhead
Weights = 320.3B × 4.8 bits ÷ 8 = 192.2 GB
KV Cache + Overhead ≈ 0.7 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 ≈ 94.8 GB (at full 524K context)
VRAM usage by quantization
Can NVIDIA GeForce RTX 5090 run IQuest Q1?
No — IQuest Q1 requires at least 136.8 GB at Q2_K, which exceeds the NVIDIA GeForce RTX 5090's 32 GB of VRAM.
What's the best quantization for IQuest Q1?
For IQuest Q1, Q4_K_M (192.9 GB) offers the best balance of quality and VRAM usage. Q5_K_M (228.9 GB) provides better quality if you have the VRAM. The smallest option is Q2_K at 136.8 GB.
VRAM requirement by quantization
★ Recommended — best balance of quality and VRAM usage.
Can I run IQuest Q1 on a Mac?
Yes — Mac Pro M2 Ultra (192 GB) and 5 other Macs can run IQuest Q1, 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 192.9 GB of usable unified memory (RAM minus macOS overhead), from Mac Studio M5 Ultra (256 GB).
Can I run IQuest Q1 locally?
Yes — IQuest Q1 can run locally on consumer hardware. Popular tools include Ollama, LM Studio, and llama.cpp.
How fast is IQuest Q1?
At Q4_K_M, IQuest Q1 can reach ~51 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 B300 → 9.0 GB active ÷ (8000 × 0.65) = 1.72 ms, plus 88 layers × 0.055 ms = 4.84 ms, so 1000 ÷ 6.56 ms = ~152 tok/s
Estimated speed at Q4_K_M (192.9 GB)
Real-world results typically within ±20%. Speed depends on batch size, quantization kernel, and software stack.
What's the download size of IQuest Q1?
At Q4_K_M, the download is about 192.19 GB. The full-precision BF16 version is 640.64 GB. The smallest option (Q2_K) is 136.14 GB.
Which GPUs can run IQuest Q1?
No single consumer GPU has enough VRAM to run IQuest Q1 at Q4_K_M (192.9 GB). Multi-GPU or professional hardware is required.
Which devices can run IQuest Q1?
4 devices with unified memory can run IQuest Q1 at Q4_K_M (192.9 GB), including Mac Studio (M3 Ultra, 256GB), Mac Studio (M3 Ultra, 512GB), Mac Studio M5 Ultra (256 GB), Mac Studio M5 Ultra (512 GB). Apple Silicon Macs use unified memory shared between CPU and GPU, making them well-suited for local LLM inference.