DBRX Instruct
By Databricks · United States
Updated 2026-07-13
Overview
Databricks' 132B MoE with 36B active params, trained on 12T tokens — state-of-the-art at March 2024 release but largely surpassed by DeepSeek V3 and R1.
When to pick this model
- Databricks-native pipelines that need an in-house model
- Code and math workloads where 36B active params shine
- Research comparisons against modern frontier MoEs
- Multi-GPU deployments already provisioned for 100B+ models
- Internal evals before migrating to DeepSeek V3 or R1
VRAM requirements by quantization
| Quantization | VRAM required |
|---|---|
| Q4_K_M (recommended) | 76 GB |
| Q5_K_M | 94 GB |
| Q8_0 | 140 GB |
| FP16 (no quantization) | 264 GB |
VRAM figures include model weights plus a typical 8k KV cache and ~600 MB runtime overhead (Ollama / llama.cpp baseline). Add headroom for higher context lengths.
In practice, DBRX Instruct is server-class even at Q4_K_M (76 GB). Stepping up to Q8_0 nearly doubles the footprint to 140 GB, and unquantized FP16 weights take 264 GB — plan your GPU around the Q4 or Q5 figure unless you specifically need the higher fidelity.
Without a GPU, DBRX Instruct needs roughly 112 GB of system RAM to run on CPU via llama.cpp or Ollama — workable for background jobs, but far slower than GPU inference. Throughput estimates from our compatibility engine: around 2 tokens/sec on entry-level GPUs, on the order of 8 tokens/sec on a mid-range card, and up to 22 tokens/sec on high-end hardware — assuming the chosen quantization fully fits in VRAM.
What hardware do you need
The table below matches DBRX Instruct to common GPU memory tiers, using the highest-fidelity quantization that fully fits each card class. Spilling layers to system RAM works but costs most of the speed, so size your card to the quantization you actually want to run.
| GPU memory | Example cards | Best fit for DBRX Instruct |
|---|---|---|
| 8 GB | RTX 5070 Laptop, RTX 5060, RTX 5060 Ti 8GB | Does not fit — needs 76 GB at Q4_K_M |
| 12 GB | RTX 5070, RTX 5070 Ti Laptop, RTX 4080 Laptop | Does not fit — needs 76 GB at Q4_K_M |
| 16 GB | RTX 5080, RTX 4080 Super, Radeon RX 9070 XT | Does not fit — needs 76 GB at Q4_K_M |
| 24 GB | RTX 4090, Radeon RX 7900 XTX, RTX 5090 Laptop | Does not fit — needs 76 GB at Q4_K_M |
| 32 GB | RTX 5090 | Does not fit — needs 76 GB at Q4_K_M |
Which GPU should you buy to run DBRX Instruct?
To run DBRX Instruct locally at Q4, you need ~76 GB of VRAM. The best value for this is a Apple Mac Studio (64+ GB unified memory).
As an Amazon Associate, BestLLMfor earns from qualifying purchases, at no extra cost to you. It does not influence our independent rankings.
Published benchmark scores
| Benchmark | Score |
|---|---|
| MMLU | 73.7 |
| HumanEval | 70.1 |
Scores published by the model author or aggregated from public leaderboards. Re-measured monthly by our editorial team.
To put DBRX Instruct in context: its MMLU score of 73.7 ranks #17 of the 34 catalog models with a published MMLU result (catalog median 73.4); its HumanEval score of 70.1 ranks #20 of the 26 catalog models with a published HumanEval result (catalog median 81.1). Scores are author-published and measured under different harnesses, so treat gaps of a few points as noise rather than a verdict.
Strengths
- State-of-the-art quality at March 2024 release
- Strong on code and math benchmarks
- Databricks Open Model License is broadly permissive
- 12T tokens of high-quality training data
Limitations
- ~76 GB VRAM at Q4 demands multi-GPU serving
- Largely outclassed by DeepSeek V3 and R1 in 2025
- HuggingFace repo is gated, slowing access
Typical workloads
In our catalog grid, DBRX Instruct is filed under Enterprise MoE, Pro Chat — the use cases where its size/quality trade-off makes the most sense. Its tags translate to concrete workloads: everyday chat, drafting and summarization.
The 32k-token context window covers long chats and mid-sized documents, though very large retrieval workloads will need chunking. It ships under the Databricks Open Model License license — commercial use is generally possible but read the specific terms before embedding it in a product.
Architecture & training
Architecture: MoE · 132B total / 36B active · 16 experts, 4 active per token
Training: Databricks — 12T high-quality tokens, strong in code and science.
Historically important but no longer competitive — only choose it inside Databricks pipelines where the integration justifies the cost.
Quick start
ollama run dbrxOr use the open-source MCP server to query this model from Claude Desktop, Cursor, or any MCP-compatible client.
Similar models worth comparing
Frequently asked questions
How much VRAM does DBRX Instruct need?
At the recommended Q4_K_M quantization, DBRX Instruct needs about 76 GB of VRAM. Q8_0 takes 140 GB, and unquantized FP16 weights take 264 GB.
Can DBRX Instruct run without a GPU?
Yes — with roughly 112 GB of system RAM it runs CPU-only through llama.cpp or Ollama. Expect a fraction of GPU speed, which is fine for background or batch jobs but slow for interactive chat.
What context window does DBRX Instruct support?
DBRX Instruct supports a 32k-token context window (32,768 tokens).
Can I use DBRX Instruct commercially?
DBRX Instruct ships under the Databricks Open Model License license. Commercial use is generally permitted subject to its terms — review the license text before shipping a product.
How fast is DBRX Instruct on consumer hardware?
Our compatibility engine estimates on the order of 8 tokens/sec on a mid-range GPU and up to 22 tokens/sec on high-end cards, assuming the quantization fully fits in VRAM.
Which quantization of DBRX Instruct should I download first?
Start with Q4_K_M (76 GB) — the standard size/quality sweet spot. Step up to Q5_K_M or Q8_0 only if you have VRAM headroom. It does not fit a single 24 GB consumer card — plan for multi-GPU or server hardware.