Salamandra 7B Instruct — Hardware Requirements & GPU Compatibility
Chat
Salamandra 7B Instruct is a 7.8-billion-parameter multilingual model developed by the Barcelona Supercomputing Center (BSC-LT) as part of a European initiative to build high-quality open language models. It has particular strength in Iberian languages including Spanish, Catalan, Portuguese, and Basque, while also supporting English and other major European languages. This model is an excellent choice for users who need strong performance in Spanish or other Iberian languages that are often underserved by mainstream LLMs. Running it locally ensures data privacy for sensitive multilingual workflows, and at 7B parameters it fits comfortably on a single consumer GPU with 8 GB or more of VRAM.
Runs great on a 24 GB GPU
Runs from 8 GB (Q4_K_M · 8K)Best on 24 GB (Q4_K_M · ~125 tok/s · 8K)Macs from 16 GB (Q4_K_M · ~27 tok/s)
Apache 2.0
Reference GPU · RTX 4090 24 GB
Estimates, ±20% · how we estimate
Run Salamandra 7B Instruct
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/RichardErkhov/BSC-LT_-_salamandra-7b-instruct-gguf:Q4_K_MQ4_K_M fits fully on the RTX 4090 · 18.0 GB free · 8K context
Download: salamandra-7b-instruct.Q4_K_M.gguf4.85 GBRichardErkhov
Ways to runOther ways
each opens step-by-stepRun in cloud
Your own private copy · Q4_K_M · 24 GB GPU
- The easiest way to try this model
- Running in a few minutes, easy setup
- Chat right here or connect any app
$5+ bonus for new users
Runs on RTX 3080 Ti (12 GB) (6 GB headroom) · Q4_K_M
- Generation speed
- ~113 tok/s
- generation speed
- Cost per 1M output tokens
- $0.30
- per 1M output tokens
How Much VRAM Does Salamandra 7B Instruct Need?
Select a quantization level; the listed hardware adheres to it.
| Quantization | VRAM | +32K | On RTX 4090 (reference) | |
|---|---|---|---|---|
| 3.9 GB | — | 2-bit · K-quant — 2-bit quantization with K-quant improvements | Fits: 8K context · ~169 tok/s | |
| 4.0 GB | — | 3-bit small — 3-bit small quantization | Fits: 8K context · ~165 tok/s | |
| 4.4 GB | — | 3-bit medium — 3-bit medium quantization | Fits: 8K context · ~150 tok/s | |
| 4.5 GB | — | 4-bit · legacy — 4-bit legacy quantization | Fits: 8K context · ~147 tok/s | |
| 5.2 GB | — | 4-bit medium — 4-bit medium quantization — most popular sweet spot | Fits: 8K context · ~125 tok/s | |
| 6.1 GB | — | 5-bit medium — 5-bit medium quantization — good quality/size tradeoff | Fits: 8K context · ~107 tok/s | |
| 7.0 GB | — | 6-bit, very good — 6-bit quantization, very good quality | Fits: 8K context · ~94 tok/s | |
| 8.3 GB | — | 8-bit, near-lossless — 8-bit quantization, near-lossless | Fits: 8K context · ~79 tok/s |
Selected Q4_K_M
4-bit medium quantization6.0GBat 8K context
- Weights
- 4.7 GB
- KV cache
- 1.1 GB
- Overhead
- 0.3 GB
Fits RTX 4090 (reference) · 18.0 GB free · ~125 tok/s
≥50 tok/s needs ≥402 GB/s · RTX 5090 ~223, RTX 3090 Ti ~125
Which Hardware Can Run Salamandra 7B Instruct?
Q4_K_M · 5.2 GBSalamandra 7B Instruct (Q4_K_M) requires 5.2 GB of VRAM to load the model weights. For comfortable inference with headroom for KV cache and system overhead, 7+ GB is recommended. Using the full 8K context window can add up to 0.8 GB, bringing total usage to 6.0 GB. 52 GPUs can run it, including NVIDIA GeForce RTX 5090, NVIDIA GeForce RTX 3090 Ti. 151 devices with unified memory can run Salamandra 7B Instruct, including Mac Studio (M3 Ultra, 256GB), Mac Studio (M3 Ultra, 512GB).
≤ 11 GB
≤ 11 GB tier →fits · 2.0 GB free · up to 8K ctx
16 GPUs · 1 device family
RTX 3080~95Simulate output on RTX 3080RTX 2080 Ti~77Simulate output on RTX 2080 TiRTX 3070 Ti~76Simulate output on RTX 3070 TiOrin Nano~13Simulate output on NVIDIA Jetson Orin Nano 8GB (Super)20 GB
fits · 14.0 GB free · up to 8K ctx
1 GPU
RX 7900 XT~92Simulate output on RX 7900 XT32–40 GB
fits · 26.0 GB free · up to 8K ctx
1 GPU · 3 Mac classes · 4 device families
RTX 5090~223Simulate output on RTX 5090M5 Max36 GB~62Simulate output on Mac Studio M5 Max (36 GB)M532 GB~21Simulate output on MacBook Air 13" M5 (32 GB)Orin~26Simulate output on NVIDIA Jetson AGX Orin 32GBSnapdragon X Elite (X1E-80-100)~17Simulate output on Snapdragon X Elite Copilot+ PC
By VRAM tier · typical GPUs and Macs at Q4_K_M
Similar models
6| Model | Score | VRAM | tok/s | Why it is similar |
|---|---|---|---|---|
| Ling 3.0 TinyInclusion AI | — | 5.5 GB | ~364 | same size |
| BitCPM CANN 8BOpenBMB | — | 5.6 GB | ~127 | same size |
| Gemma 4 E4BGoogle | — | 5.6 GB | ~123 | same size |
| Olmo 3 7B InstructAllen AI | — | 8.4 GB | ~114 | same size |
| Phi Mini MoE InstructMicrosoft | — | 5.4 GB | ~239 | same size |
| Gemma 3n E4B ITGoogle | — | 7.7 GB | ~127 | same size |
Specifications
- Publisher
- BSC-LT
- Parameters
- 7.8B
- Architecture
- LlamaForCausalLM
- Context Length
- 8,192 tokens
- Vocabulary Size
- 256,000
- Release Date
- 2024-09-30
- License
- Apache 2.0
Frequently Asked Questions
How much VRAM does Salamandra 7B Instruct need?
Salamandra 7B Instruct requires 5.2 GB of VRAM at Q4_K_M, or 16.1 GB at BF16. Full 8K context adds up to 0.8 GB (6.0 GB total).
VRAM = Weights + KV Cache + Overhead
Weights = 7.8B × 4.8 bits ÷ 8 = 4.7 GB
KV Cache + Overhead ≈ 0.5 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 ≈ 1.3 GB (at full 8K context)
VRAM usage by quantization
What's the best quantization for Salamandra 7B Instruct?
For Salamandra 7B Instruct, Q4_K_M (5.2 GB) offers the best balance of quality and VRAM usage. Q5_0 (5.4 GB) provides better quality if you have the VRAM. The smallest option is IQ3_XS at 3.8 GB.
VRAM requirement by quantization
★ Recommended — best balance of quality and VRAM usage.
Can I run Salamandra 7B Instruct on a Mac?
Yes — Mac Mini M2 Pro (16 GB) and 133 other Macs can run Salamandra 7B Instruct. Apple Silicon uses unified memory, so the model shares RAM with the system. At Q4_K_M you need at least 5.2 GB of usable unified memory (RAM minus macOS overhead).
Can I run Salamandra 7B Instruct locally?
Yes — Salamandra 7B Instruct can run locally on consumer hardware. Popular tools include Ollama, LM Studio, and llama.cpp.
How fast is Salamandra 7B Instruct?
At Q4_K_M, Salamandra 7B Instruct can reach ~918 tok/s on AMD Instinct MI350X. On NVIDIA GeForce RTX 4090: ~125 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 ÷ 5.23 × 0.65 = ~994 tok/s
Estimated speed at Q4_K_M (5.2 GB)
Real-world results typically within ±20%. Speed depends on batch size, quantization kernel, and software stack.
What's the download size of Salamandra 7B Instruct?
At Q4_K_M, the download is about 4.66 GB. The full-precision BF16 version is 15.54 GB. The smallest option (IQ3_XS) is 3.20 GB.
Which GPUs can run Salamandra 7B Instruct?
52 consumer GPUs can run Salamandra 7B Instruct 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 Salamandra 7B Instruct?
151 devices with unified memory can run Salamandra 7B Instruct at Q4_K_M (5.2 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.