T5 Paraphrase Paws — Hardware Requirements & GPU Compatibility
Chat
T5 Paraphrase Paws is a 223M-parameter open language model from Vamsi. It supports a context window of up to 512 tokens. At Q4_K_M it needs about 0.15 GB of VRAM — see which GPUs and Macs can run it below.
Runs great on a 24 GB GPU
Runs from 8 GB (Q4_K_M)Best on 24 GB (Q4_K_M · ~4368 tok/s)Macs from 8 GB (Q4_K_M · ~478 tok/s)
Reference GPU · RTX 4090 24 GB
Estimates, ±20% · how we estimate
Run T5 Paraphrase Paws
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/tensorblock/T5_Paraphrase_Paws-GGUFQ4_K_M fits fully on the RTX 4090 · 23.8 GB free · 512 context
Download: T5_Paraphrase_Paws-Q4_K_M.gguf0.13 GBtensorblock
Ways to runOther ways
each opens step-by-stepRun in cloud
Runs on RTX 3080 Ti (12 GB) (11 GB headroom) · Q4_K_M
- Generation speed
- ~3954 tok/s
- generation speed
- Cost per 1M output tokens
- $0.0086
- per 1M output tokens
How Much VRAM Does T5 Paraphrase Paws Need?
Select a quantization level; the listed hardware adheres to it.
| Quantization | VRAM | +32K | On RTX 4090 (reference) | |
|---|---|---|---|---|
| 0.1 GB | — | 2-bit · K-quant — 2-bit quantization with K-quant improvements | Fits: — context · ~6552 tok/s | |
| 0.1 GB | — | 3-bit medium — 3-bit medium quantization | Fits: — context · ~5460 tok/s | |
| 0.1 GB | — | 4-bit medium — 4-bit medium quantization — most popular sweet spot | Fits: — context · ~4368 tok/s | |
| 0.2 GB | — | 5-bit medium — 5-bit medium quantization — good quality/size tradeoff | Fits: — context · ~3854 tok/s | |
| 0.2 GB | — | 6-bit, very good — 6-bit quantization, very good quality | Fits: — context · ~3276 tok/s | |
| 0.3 GB | — | 8-bit, near-lossless — 8-bit quantization, near-lossless | Fits: — context · ~2621 tok/s | |
| 0.5 GB | — | 16-bit · bfloat — Brain floating point 16 — preferred for training | Fits: — context · ~1337 tok/s |
Selected Q4_K_M
4-bit medium quantization0.2GBat 1K context
- Weights
- 0.1 GB
- KV cache
- 0.0 GB
- Overhead
- 0.3 GB
Fits RTX 4090 (reference) · 23.9 GB free · ~4368 tok/s
≥50 tok/s needs ≥12 GB/s · RTX 5090 ~7765, RTX 3090 Ti ~4368
Which Hardware Can Run T5 Paraphrase Paws?
Q4_K_M · 0.1 GBT5 Paraphrase Paws (Q4_K_M) requires 0.1 GB of VRAM to load the model weights. For comfortable inference with headroom for KV cache and system overhead, 1+ GB is recommended. 52 GPUs can run it, including NVIDIA GeForce RTX 5090, NVIDIA GeForce RTX 3090 Ti. 163 devices with unified memory can run T5 Paraphrase Paws, including Mac Studio (M3 Ultra, 256GB), Mac Studio (M3 Ultra, 512GB).
≤ 11 GB
≤ 11 GB tier →fits · 7.8 GB free · up to — ctx
16 GPUs · 1 Mac class · 2 device families
RTX 3080~3295Simulate output on RTX 3080RTX 2080 Ti~2669Simulate output on RTX 2080 TiRTX 3070 Ti~2636Simulate output on RTX 3070 TiM3*8 GB~478Simulate output on iMac 24" M3 (8 GB)Orin Nano~442Simulate output on NVIDIA Jetson Orin Nano 8GB (Super)iPhone~318Simulate output on iPhone 1720 GB
fits · 19.9 GB free · up to — ctx
1 GPU
RX 7900 XT~3200Simulate output on RX 7900 XT32–40 GB
fits · 31.9 GB free · up to — ctx
1 GPU · 3 Mac classes · 4 device families
RTX 5090~7765Simulate output on RTX 5090M5 Max36 GB~2147Simulate output on Mac Studio M5 Max (36 GB)M532 GB~717Simulate output on MacBook Air 13" M5 (32 GB)Orin~888Simulate output on NVIDIA Jetson AGX Orin 32GBSnapdragon X Elite (X1E-80-100)~585Simulate output on Snapdragon X Elite Copilot+ PC
By VRAM tier · typical GPUs and Macs at Q4_K_M
Similar models
2| Model | Score | VRAM | tok/s | Why it is similar |
|---|---|---|---|---|
| LFM2.5 230MLiquid AI | — | 0.6 GB | ~1310 | same size |
| MiniCPM V 4.7 1B DSparkOpenBMB | — | 0.5 GB | ~1489 | same size |
Specifications
- Publisher
- Vamsi
- Parameters
- 223M
- Architecture
- T5ForConditionalGeneration
- Context Length
- 512 tokens
- Vocabulary Size
- 32,128
- Release Date
- 2022-03-02
Frequently Asked Questions
How much VRAM does T5 Paraphrase Paws need?
T5 Paraphrase Paws requires 0.1 GB of VRAM at Q4_K_M, or 0.5 GB at BF16.
VRAM = Weights + KV Cache + Overhead
Weights = 223M × 4.8 bits ÷ 8 = 0.1 GB
Fit ratings and hardware model lists check this model with room for a 16K-token context, which needs a little more memory.
VRAM usage by quantization
What's the best quantization for T5 Paraphrase Paws?
For T5 Paraphrase Paws, Q4_K_M (0.1 GB) offers the best balance of quality and VRAM usage. Q5_K_M (0.2 GB) provides better quality if you have the VRAM. The smallest option is Q2_K at 0.1 GB.
VRAM requirement by quantization
★ Recommended — best balance of quality and VRAM usage.
Can I run T5 Paraphrase Paws on a Mac?
Yes — MacBook Air 13" M3 (8 GB) and 145 other Macs can run T5 Paraphrase Paws. Apple Silicon uses unified memory, so the model shares RAM with the system. At Q4_K_M you need at least 0.1 GB of usable unified memory (RAM minus macOS overhead).
Can I run T5 Paraphrase Paws locally?
Yes — T5 Paraphrase Paws can run locally on consumer hardware. Popular tools include Ollama, LM Studio, and llama.cpp.
How fast is T5 Paraphrase Paws?
At Q4_K_M, T5 Paraphrase Paws can reach ~32000 tok/s on AMD Instinct MI350X. On NVIDIA GeForce RTX 4090: ~4368 tok/s. Speed depends mainly on GPU memory bandwidth. Real-world results typically within ±20%.
tok/s = (bandwidth GB/s ÷ model GB) × efficiency
Example: NVIDIA B200 → 8000 ÷ 0.15 × 0.65 = ~34667 tok/s
Estimated speed at Q4_K_M (0.1 GB)
Real-world results typically within ±20%. Speed depends on batch size, quantization kernel, and software stack.
What's the download size of T5 Paraphrase Paws?
At Q4_K_M, the download is about 0.13 GB. The full-precision BF16 version is 0.45 GB. The smallest option (Q2_K) is 0.09 GB.
Which GPUs can run T5 Paraphrase Paws?
52 consumer GPUs can run T5 Paraphrase Paws at Q4_K_M. Top options include AMD Radeon RX 6700 XT, AMD Radeon RX 6800, AMD Radeon RX 6800 XT. 52 GPUs have plenty of headroom for comfortable inference.
Which devices can run T5 Paraphrase Paws?
163 devices with unified memory can run T5 Paraphrase Paws at Q4_K_M (0.1 GB), including AMD Ryzen AI 9 HX 370 (Strix Point) Laptop, ASUS Ascent GX10, Beelink GTR9 Pro (Ryzen AI Max+ 395, 128 GB), Framework Desktop (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.