Starcoder2 7B — Hardware Requirements & GPU Compatibility
ChatCode
Starcoder2 7B is a 7.2B-parameter open language model from BigCode in the StarCoder family. It supports a context window of up to 16,384 tokens. At Q4_K_M it needs about 4.74 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 · 16K)Best on 24 GB (Q4_K_M · ~138 tok/s · 16K)Macs from 8 GB (Q2_K · ~21 tok/s)
bigcode-openrail-m
Reference GPU · RTX 4090 24 GB
Estimates, ±20% · how we estimate
Run Starcoder2 7B
One recommended way. Every other method, OS and quant is in the step-by-step guide.
Recommended
ollama run starcoder2:7b-q4_K_MQ4_K_M fits fully on the RTX 4090 · 18.3 GB free · 16K context
Download: starcoder2-7b-Q4_K_M.gguf4.40 GBsecond-state
Other ways
each opens step-by-stepRun in cloud
Fits on RTX 3060 12GB (6 GB headroom) · Q4_K_M
- Generation speed
- ~49 tok/s
- generation speed
- Cost per 1M output tokens
- $0.47
- per 1M output tokens
How Much VRAM Does Starcoder2 7B Need?
Select a quantization level; the listed hardware adheres to it.
| Quantization | VRAM | +32K | On RTX 4090 (reference) | |
|---|---|---|---|---|
| 3.5 GB | — | 2-bit · K-quant — 2-bit quantization with K-quant improvements | Fits: 16K context · ~188 tok/s | |
| 3.6 GB | — | 3-bit small — 3-bit small quantization | Fits: 16K context · ~184 tok/s | |
| 3.9 GB | — | 3-bit medium — 3-bit medium quantization | Fits: 16K context · ~167 tok/s | |
| 4.0 GB | — | 4-bit · legacy — 4-bit legacy quantization | Fits: 16K context · ~163 tok/s | |
| 4.7 GB | — | 4-bit medium — 4-bit medium quantization — most popular sweet spot | Fits: 16K context · ~138 tok/s | |
| 5.5 GB | — | 5-bit medium — 5-bit medium quantization — good quality/size tradeoff | Fits: 16K context · ~118 tok/s | |
| 6.3 GB | — | 6-bit, very good — 6-bit quantization, very good quality | Fits: 16K context · ~103 tok/s | |
| 7.6 GB | — | 8-bit, near-lossless — 8-bit quantization, near-lossless | Fits: 16K context · ~86 tok/s |
Selected Q4_K_M
4-bit medium quantization5.7GBat 16K context
- Weights
- 4.3 GB
- KV cache
- 1.1 GB
- Overhead
- 0.3 GB
Fits RTX 4090 (reference) · 18.3 GB free · ~138 tok/s
≥50 tok/s needs ≥364 GB/s · RTX 5090 ~246, RTX 3090 Ti ~138
Which Hardware Can Run Starcoder2 7B?
Q4_K_M · 4.7 GBStarcoder2 7B (Q4_K_M) requires 4.7 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 16K context window can add up to 0.9 GB, bringing total usage to 5.7 GB. 52 GPUs can run it, including NVIDIA GeForce RTX 5090, NVIDIA GeForce RTX 3090 Ti. 151 devices with unified memory can run Starcoder2 7B, including Mac Studio (M3 Ultra, 256GB), Mac Studio (M3 Ultra, 512GB).
≤ 11 GB
≤ 11 GB tier →fits · 2.3 GB free · up to 16K ctx
16 GPUs · 2 device families
RTX 3080~104Simulate output on RTX 3080RTX 2080 Ti~85Simulate output on RTX 2080 TiRTX 3070 Ti~83Simulate output on RTX 3070 TiOrin Nano~14Simulate output on NVIDIA Jetson Orin Nano 8GB (Super)iPhone~10Simulate output on iPhone 1720 GB
fits · 14.3 GB free · up to 16K ctx
1 GPU
RX 7900 XT~101Simulate output on RX 7900 XT32–40 GB
fits · 26.3 GB free · up to 16K ctx
1 GPU · 3 Mac classes · 4 device families
RTX 5090~246Simulate output on RTX 5090M5 Max36 GB~68Simulate output on Mac Studio M5 Max (36 GB)M532 GB~23Simulate output on MacBook Air 13" M5 (32 GB)Orin~28Simulate output on NVIDIA Jetson AGX Orin 32GBSnapdragon X Elite (X1E-80-100)~19Simulate output on Snapdragon X Elite Copilot+ PC
By VRAM tier · typical GPUs and Macs at Q4_K_M
How good is Starcoder2 7B?
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
Here are all 4 boards that have results for this model, sorted by their weight in the llmrun Score; badges show who ran the test.
- MMLU
Epoch AI · Feb 2024Lab-reported
Lab-reported38.8%
#72/85 open
- GSM8K
Epoch AI · Feb 2024Lab-reported
Lab-reported32.7%
#44/62 open
- ARC Challenge
Epoch AI · Feb 2024Lab-reported
Lab-reported38.7%
#43/56 open
- WinoGrande
Epoch AI · Feb 2024Lab-reported
Lab-reported57.1%
#47/51 open
Alternatives to Starcoder2 7B
StarCoder family · 5 models →Similar models
6| Model | Score | VRAM | tok/s | Why it is similar |
|---|---|---|---|---|
| Qwen2.5 7B InstructAlibaba | 65.1 | 5.8 GB | ~131 | same size |
| Olmo 3 7B InstructAllen AI | — | 8.4 GB | ~114 | same size |
| Phi Mini MoE InstructMicrosoft | — | 5.4 GB | ~239 | same size |
| Falcon H1 7B InstructTII UAE | — | 6.3 GB | ~130 | same size |
| DeepSeek R1 Distill Qwen 7BDeepSeek | — | 5.8 GB | ~131 | same size |
| Eurus 2 7B PRIMEPRIME-RL | — | 5.1 GB | ~131 | same size |
Specifications
- Publisher
- BigCode
- Family
- StarCoder
- Parameters
- 7.2B
- Architecture
- Starcoder2ForCausalLM
- Context Length
- 16,384 tokens
- Vocabulary Size
- 49,152
- Release Date
- 2024-02-20
- License
- bigcode-openrail-m
Frequently Asked Questions
How much VRAM does Starcoder2 7B need?
Starcoder2 7B requires 4.7 GB of VRAM at Q4_K_M, or 14.8 GB at BF16. Full 16K context adds up to 0.9 GB (5.7 GB total).
VRAM = Weights + KV Cache + Overhead
Weights = 7.2B × 4.8 bits ÷ 8 = 4.3 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 ≈ 1.4 GB (at full 16K context)
VRAM usage by quantization
What's the best quantization for Starcoder2 7B?
For Starcoder2 7B, Q4_K_M (4.7 GB) offers the best balance of quality and VRAM usage. Q5_0 (4.9 GB) provides better quality if you have the VRAM. The smallest option is Q2_K at 3.5 GB.
VRAM requirement by quantization
★ Recommended — best balance of quality and VRAM usage.
Can I run Starcoder2 7B on a Mac?
Yes — MacBook Air 13" M3 (8 GB) and 145 other Macs can run Starcoder2 7B, 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 4.7 GB of usable unified memory (RAM minus macOS overhead), from Mac Mini M2 Pro (16 GB).
Can I run Starcoder2 7B locally?
Yes — Starcoder2 7B can run locally on consumer hardware. Popular tools include Ollama, LM Studio, and llama.cpp.
How fast is Starcoder2 7B?
At Q4_K_M, Starcoder2 7B can reach ~1013 tok/s on AMD Instinct MI350X. On NVIDIA GeForce RTX 4090: ~138 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 ÷ 4.74 × 0.65 = ~1097 tok/s
Estimated speed at Q4_K_M (4.7 GB)
Real-world results typically within ±20%. Speed depends on batch size, quantization kernel, and software stack.
What's the download size of Starcoder2 7B?
At Q4_K_M, the download is about 4.30 GB. The full-precision BF16 version is 14.35 GB. The smallest option (Q2_K) is 3.05 GB.
Which GPUs can run Starcoder2 7B?
52 consumer GPUs can run Starcoder2 7B 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 Starcoder2 7B?
151 devices with unified memory can run Starcoder2 7B at Q4_K_M (4.7 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.