Databricks' pioneering 132B open MoE model utilizing 16 fine-grained experts (4 active) for enterprise SQL, coding, and general tasks.
Technical Specifications
- Provider
- Databricks
- Category
- open-source
- Release Date
- 2024-03
- Batch API Discount
- Standard rate
- Consumer Subscription
- Free / Self-Hosted
- Reasoning Effort Control
- Fixed
- Structured Outputs
- Yes
- Output Modalities
- text
- Official Documentation
- Visit Official Docs ↗
Data Privacy, Licensing & Deployment Governance
Official Provider Release Intelligence, API Quotas & Agentic Specs
- Official API Model ID (SDK String)
- dbrx-instruct
- Supported SDK & API Protocol
- OpenAI-Compatible REST / vLLM / SGLang / HuggingFace TGI
- Lifecycle & Deprecation Status
- Legacy Stable (Upgrade path available in Databricks lineup)
- Production Rate Limits (RPM & TPM)
- Unlimited (Hardware-bound on self-hosted GPU) / 10,000 RPM Cloud · Unlimited local throughput / 10M+ TPM Managed Cloud
- Long-Context Pricing Tier (>128K / >200K)
- Free across full context window (only KV-cache VRAM scales)
- Prompt Cache Write & TTL Policy
- Automatic prefix KV-caching in vLLM / SGLang (0 cost)
- Computer Use & Browser GUI Agent
- API / CLI Tool-Use & DOM Agent ready (no raw pixel desktop click loop)
- MCP & Parallel Tool Calling
- Native MCP Client/Server ready · Parallel function calling · 100% Strict JSON Schema
- Web Search Grounding & Code Sandbox
- Supported via external Tavily/Exa Search & E2B/Docker Sandbox tools
- Fine-Tuning & Distillation Support
- Full Weights SFT, LoRA, QLoRA, DPO & GRPO RL (Unsloth / Axolotl / Llama-Factory)
- Multimodal Tokenization Billing
- Text BPE (~4 chars/token)
- Weights Precision & Quantization
- BF16 / FP16, FP8 Dynamic, INT4 AWQ / GPTQ, GGUF (Q4_K_M / Q8_0)
- Minimum GPU VRAM Requirement
- 82 GB VRAM (4-bit Quantized) · 284 GB VRAM (Native BF16)
- Supported Inference Engines
- vLLM, SGLang, Ollama, llama.cpp, TensorRT-LLM, HuggingFace TGI
- System Card & Safety Framework
- Open-Weights System Card + Customizable Guardrail Layer (Llama Guard / ShieldGemma)
- Enterprise Compliance (SOC2 / HIPAA / ZDR)
- 100% Air-Gapped On-Premises / Private VPC · Full HIPAA, GDPR & SOC2 Control
DBRX Instruct Cost, Context, Hardware & API Integration
Open-weights or free-tier model: $0 API license cost when self-hosted.
Same Monthly Workload Cost Across Popular Reference Models
Live Prompt Paste Tokenizer & Cost Counter
Paste real system instructions, JSON schemas, or code to measure BPE token counts across tokenizers.
Hybrid Prompt Routing Cascade Simulator
Route routine queries to a fast lightweight model ($0.15/1M) and escalate only complex reasoning turns to DBRX Instruct:
Benchmark Scores & 6-Axis Capability Radar
- MMLU-Pro
- 74%
- HumanEval
- 78.5%
- Chatbot Arena Elo
- 1265pts
Strengths
- Pioneered fine-grained 16-expert MoE architecture in commercial open-source models
- Strong out-of-the-box SQL generation and tabular database query capabilities
- Tight native integration with the Databricks Lakehouse platform
Limitations & Trade-offs
- 32K context window is smaller than modern 128K+ alternatives
- Requires substantial GPU VRAM for unquantized cluster hosting
Editorial Assessment
A milestone open enterprise MoE that proved the efficiency of fine-grained expert routing.
Best for: Enterprise data warehouse analytics, automated SQL generation, and lakehouse assistants
Compare DBRX Instruct
Frequently Asked Questions about DBRX Instruct
How much does the DBRX Instruct API cost per 1M tokens?+
DBRX Instruct by Databricks is priced at $0 per 1M input tokens and $0 per 1M output tokens. With prompt caching enabled, cached input tokens cost $0 per 1M tokens, and asynchronous Batch API requests receive a 0% discount.
Should I use a $20/month subscription or the DBRX Instruct pay-as-you-go API?+
A flat $20/month consumer subscription is predictable for daily interactive chat and manual coding. However, if you are building software, running automated workflows, or sending low-to-moderate volume with prompt caching enabled, the pay-as-you-go API is typically far more cost-effective and offers higher rate limits without manual UI caps.
How many words and pages fit in DBRX Instruct's context window?+
DBRX Instruct features a 33K tokens context window, which equals approximately 24,576 English words or ~66 standard single-spaced pages in a single prompt. Its maximum output generation per turn is 4,096 tokens.
Can I run DBRX Instruct locally on my own GPU (Open Weights)?+
Yes, DBRX Instruct (132B total (36B active MoE)) releases open weights that can be self-hosted locally or on private cloud GPUs using vLLM, Ollama, SGLang, or llama.cpp.
Does Databricks train on data sent to the DBRX Instruct API, and is commercial use allowed?+
100% Isolated when self-hosted on private GPUs or VPC. License tier: Open-Weights Research & Commercial License. Full commercial ownership of generated outputs. Supported deployment channels include: Self-Hosted GPU (vLLM / Ollama), Private VPC, or Managed Serverless API.
How does DBRX Instruct perform on coding and reasoning benchmarks?+
DBRX Instruct achieves a composite Intelligence Index of 76/100, scoring N/A% on SWE-bench Verified, N/A% on GPQA Diamond, 74% on MMLU-Pro, and 1265 pts on Chatbot Arena Elo.
What is DBRX Instruct best used for?+
DBRX Instruct is best suited for: Enterprise data warehouse analytics, automated SQL generation, and lakehouse assistants. Assessment: A milestone open enterprise MoE that proved the efficiency of fine-grained expert routing.
AI Glossary & Architecture Concepts
View full glossary →Key machine learning concepts and benchmarks relevant to DBRX Instruct:
The basic unit of text that a language model processes, typically representing about 3 to 4 English characters or roughly 0.75 words. For example, the sentence "Artificial intelligence is fast" breaks down into about 5 tokens.
A standardized evaluation suite used to objectively measure and rank an AI model's capabilities across specific tasks — such as SWE-bench Verified for software engineering, AIME for competitive math, or MMLU for general knowledge.
The maximum number of tokens a model can consider at once — both the input it reads and the output it writes. A 1,000,000-token context window can hold roughly 750,000 English words in a single request.
The ability of a language model to recognize when an external capability is needed and output structured JSON arguments to invoke external APIs, databases, or terminal commands. The runtime executes the tool and passes results back to the model.
An open standard created by Anthropic that allows AI applications and agents to connect seamlessly to local filesystems, development environments, and third-party APIs via standardized server interfaces.
An architecture where a large model is split into many specialist "expert" sub-networks and only a few activate per token. This keeps inference cheap while allowing a huge total parameter count — for example DeepSeek-V3 has 671B total parameters but activates only 37B per token.
Benchmark scores, token prices, and rate limits evolve rapidly as AI labs release model updates. Always verify current specifications and pricing on the official provider documentation.