Internlm2 5 20B Chat — Hardware Requirements & GPU Compatibility
Chat
Internlm2 5 20B Chat is a 19.9B-parameter open language model from InternLM in the InternLM family. It supports a context window of up to 32,768 tokens. At Q4_K_M it needs about 12.62 GB of VRAM — see which GPUs and Macs can run it below.
Runs well on a 24 GB GPU
Runs from 12 GB (Q2_K · 16K)Best on 24 GB (Q4_K_M · ~52 tok/s · 32K)Macs from 16 GB (Q2_K · ~15 tok/s)
Other
Reference GPU · RTX 4090 24 GB
Estimates, ±20% · how we estimate
Run Internlm2 5 20B Chat
One recommended way. Every other method, OS and quant is in the step-by-step guide.
Recommended
vllm serve internlm/internlm2_5-20b-chat --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 RTX A5000 (24 GB) (8 GB headroom) · Q4_K_M
- Generation speed
- ~40 tok/s
- generation speed
- Cost per 1M output tokens
- $1.12
- per 1M output tokens
How Much VRAM Does Internlm2 5 20B Chat Need?
Select a quantization level; the listed hardware adheres to it.
| Quantization | VRAM | +32K | On RTX 4090 (reference) | |
|---|---|---|---|---|
| 9.1 GB | 15.2 GB | 2-bit · K-quant — 2-bit quantization with K-quant improvements | Fits: 32K context · ~72 tok/s | |
| 10.4 GB | 16.4 GB | 3-bit medium — 3-bit medium quantization | Fits: 32K context · ~63 tok/s | |
| 12.6 GB | 18.7 GB | 4-bit medium — 4-bit medium quantization — most popular sweet spot | Fits: 32K context · ~52 tok/s | |
| 14.8 GB | 20.9 GB | 5-bit medium — 5-bit medium quantization — good quality/size tradeoff | Fits: 32K context · ~44 tok/s | |
| 17.1 GB | 23.1 GB | 6-bit, very good — 6-bit quantization, very good quality | Fits: 32K context · ~38 tok/s | |
| 20.6 GB | 26.6 GB | 8-bit, near-lossless — 8-bit quantization, near-lossless | Fits: 16K context · ~32 tok/s | |
| 40.4 GB | 46.5 GB | 16-bit · bfloat — Brain floating point 16 — preferred for training | Does not fit: needs 43.2 GB |
Selected Q4_K_M
4-bit medium quantization15.4GBat 16K context
- Weights
- 11.9 GB
- KV cache
- 3.2 GB
- Overhead
- 0.3 GB
Fits RTX 4090 (reference) · 8.6 GB free · ~52 tok/s
≥50 tok/s needs ≥970 GB/s · RTX 5090 ~92, RTX 3090 Ti ~52
Which Hardware Can Run Internlm2 5 20B Chat?
Q4_K_M · 12.6 GBInternlm2 5 20B Chat (Q4_K_M) requires 12.6 GB of VRAM to load the model weights. For comfortable inference with headroom for KV cache and system overhead, 17+ GB is recommended. Using the full 33K context window can add up to 6.0 GB, bringing total usage to 18.7 GB. 26 GPUs can run it, including NVIDIA GeForce RTX 5090, NVIDIA GeForce RTX 3090 Ti, NVIDIA GeForce RTX 5080. 125 devices with unified memory can run Internlm2 5 20B Chat, including Mac Studio (M3 Ultra, 256GB), Mac Studio (M3 Ultra, 512GB), MacBook Pro 14" M3 Pro (18 GB).
16–18 GB
16–18 GB tier →fits · 0.6 GB free · up to 16K ctx
18 GPUs · 1 Mac class · 1 device family
RTX 5080~49Simulate output on RTX 5080RTX 5070 Ti~46Simulate output on RTX 5070 TiRTX 4080 SUPER~38Simulate output on RTX 4080 SUPERM3 Pro*18 GB~8Simulate output on MacBook Pro 14" M3 Pro (18 GB)Orin NX~5Simulate output on NVIDIA Jetson Orin NX 16GB20 GB
fits · 4.6 GB free · up to 32K ctx
1 GPU
RX 7900 XT~38Simulate output on RX 7900 XT32–40 GB
fits · 16.6 GB free · up to 32K ctx
1 GPU · 3 Mac classes · 4 device families
RTX 5090~92Simulate output on RTX 5090M5 Max36 GB~26Simulate output on Mac Studio M5 Max (36 GB)M532 GB~9Simulate output on MacBook Air 13" M5 (32 GB)Orin~11Simulate output on NVIDIA Jetson AGX Orin 32GBSnapdragon X Elite (X1E-80-100)~7Simulate output on Snapdragon X Elite Copilot+ PC
By VRAM tier · typical GPUs and Macs at Q4_K_M
How good is Internlm2 5 20B Chat?
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
Arena1190.8
#149/187 open
Alternatives to Internlm2 5 20B Chat
InternLM family · 7 models →Similar models
4| Model | Score | VRAM | tok/s | Why it is similar |
|---|---|---|---|---|
| InternVL3 5 GPT OSS 20B A4B Preview HFOpenGVLab | — | 13.6 GB | ~180 | same size |
| Tri 21B Thinktrillionlabs | — | 45.1 GB | —needs > 24 GB | same size |
| ERNIE 4.5 21B A3B PTBaidu | — | 13.8 GB | ~254 | same size |
| GPT Neox 20BEleutherAI | — | 13.7 GB | ~48 | same size |
Specifications
- Publisher
- InternLM
- Family
- InternLM
- Parameters
- 19.9B
- Architecture
- InternLM2ForCausalLM
- Context Length
- 32,768 tokens
- Vocabulary Size
- 92,544
- Release Date
- 2024-07-30
- License
- Other
Frequently Asked Questions
How much VRAM does Internlm2 5 20B Chat need?
Internlm2 5 20B Chat requires 12.6 GB of VRAM at Q4_K_M, or 40.4 GB at BF16. Full 33K context adds up to 6.0 GB (18.7 GB total).
VRAM = Weights + KV Cache + Overhead
Weights = 19.9B × 4.8 bits ÷ 8 = 11.9 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 ≈ 6.8 GB (at full 33K context)
VRAM usage by quantization
Can NVIDIA GeForce RTX 4090 run Internlm2 5 20B Chat?
Yes, at Q8_0 (20.6 GB) or lower. Higher quantizations like BF16 (40.4 GB) exceed the NVIDIA GeForce RTX 4090's 24 GB.
What's the best quantization for Internlm2 5 20B Chat?
For Internlm2 5 20B Chat, Q4_K_M (12.6 GB) offers the best balance of quality and VRAM usage. Q5_K_M (14.8 GB) provides better quality if you have the VRAM. The smallest option is Q2_K at 9.1 GB.
VRAM requirement by quantization
★ Recommended — best balance of quality and VRAM usage.
Can I run Internlm2 5 20B Chat on a Mac?
Yes — Mac Mini M2 Pro (16 GB) and 133 other Macs can run Internlm2 5 20B Chat, the smallest at Q2_K. Apple Silicon uses unified memory, so the model shares RAM with the system. At Q4_K_M you need at least 12.6 GB of usable unified memory (RAM minus macOS overhead), from MacBook Pro 14" M3 Pro (18 GB).
Can I run Internlm2 5 20B Chat locally?
Yes — Internlm2 5 20B Chat can run locally on consumer hardware. Popular tools include Ollama, LM Studio, and llama.cpp.
How fast is Internlm2 5 20B Chat?
At Q4_K_M, Internlm2 5 20B Chat can reach ~380 tok/s on AMD Instinct MI350X. On NVIDIA GeForce RTX 4090: ~52 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 ÷ 12.62 × 0.65 = ~412 tok/s
Estimated speed at Q4_K_M (12.6 GB)
Real-world results typically within ±20%. Speed depends on batch size, quantization kernel, and software stack.
What's the download size of Internlm2 5 20B Chat?
At Q4_K_M, the download is about 11.92 GB. The full-precision BF16 version is 39.72 GB. The smallest option (Q2_K) is 8.44 GB.
Which GPUs can run Internlm2 5 20B Chat?
26 consumer GPUs can run Internlm2 5 20B Chat at Q4_K_M. Top options include AMD Radeon RX 7900 XT, AMD Radeon RX 7900 XTX, NVIDIA GeForce RTX 3090, AMD Radeon RX 6800. 8 GPUs have plenty of headroom for comfortable inference.
Which devices can run Internlm2 5 20B Chat?
125 devices with unified memory can run Internlm2 5 20B Chat at Q4_K_M (12.6 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.