de Mistral líder abierto 12-miles de millones parámetro multimodal modelo natively entrenado en texto y imágenes con arbitrary aspect ratio visión encoding.
Technical Specifications
- Provider
- Mistral AI
- Category
- open-source
- Release Date
- 2024-09
- Batch API Discount
- 50% off async batch
- 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)
- pixtral-12b-latest
- Supported SDK & API Protocol
- OpenAI-Compatible REST / vLLM / SGLang / HuggingFace TGI
- Lifecycle & Deprecation Status
- Legacy Stable (Upgrade path available in Mistral AI 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)
- Flat rate ($0.15/1M in) across full 128K window
- Prompt Cache Write & TTL Policy
- Cached input at $0.05/1M tokens
- 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) · Vision (~258–765 tokens per image tile)
- Weights Precision & Quantization
- BF16 / FP16, FP8 Dynamic, INT4 AWQ / GPTQ, GGUF (Q4_K_M / Q8_0)
- Minimum GPU VRAM Requirement
- 7 GB VRAM (4-bit Quantized) · 26 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
Pixtral 12B Cost, Context, Hardware & API Integration
Pay-as-you-go API is cheaper for this volume (saves $11.75/mo vs $20 flat plan).
Break-even point: ~1,212 requests/day equals $20/month.
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 Pixtral 12B:
Benchmark Scores & 6-Axis Capability Radar
- SWE-bench Verified
- 28%
- GPQA Diamond
- 43.8%
- MMLU-Pro
- 67.5%
- HumanEval
- 75%
- Terminal-Bench
- 24%
- AIME
- 48%
- Long-Context Reasoning
- 65%
- Chatbot Arena Elo
- 1225pts
Strengths
- nativo imagen comprensión supporting multiple imágenes y arbitrary aspect ratios
- de pesos abiertos bajo Apache 2.0 license para unrestricted local hosting
- sólido visual documento OCR, gráfico parsing, y diagram comprensión
Limitations & Trade-offs
- Smaller capacidad limits complejo multipaso razonamiento en denso schematics
- video y audio modalities no supported
Editorial Assessment
de Mistral breakthrough entry en de pesos abiertos multimodal computer visión.
Best for: documento OCR, visual inspección agentes, y abierto multimodal canalizaciones
Compare Pixtral 12B
Frequently Asked Questions about Pixtral 12B
How much does the Pixtral 12B API cost per 1M tokens?+
Pixtral 12B by Mistral AI is priced at $0.15 per 1M input tokens and $0.15 per 1M output tokens. With prompt caching enabled, cached input tokens cost $0.05 per 1M tokens, and asynchronous Batch API requests receive a 50% discount.
Should I use a $20/month subscription or the Pixtral 12B 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 Pixtral 12B's context window?+
Pixtral 12B features a 128K tokens context window, which equals approximately 96,000 English words or ~256 standard single-spaced pages in a single prompt. Its maximum output generation per turn is 8,192 tokens.
Can I run Pixtral 12B locally on my own GPU (Open Weights)?+
Yes, Pixtral 12B (12B) releases open weights that can be self-hosted locally or on private cloud GPUs using vLLM, Ollama, SGLang, or llama.cpp.
Does Mistral AI train on data sent to the Pixtral 12B API, and is commercial use allowed?+
100% Isolated when self-hosted on private GPUs or VPC. License tier: Permissive Open-Weights (Apache 2.0 / MIT Commercial). Full commercial ownership of generated outputs. Supported deployment channels include: Self-Hosted GPU (vLLM / Ollama), Private VPC, or Managed Serverless API.
How does Pixtral 12B perform on coding and reasoning benchmarks?+
Pixtral 12B achieves a composite Intelligence Index of 40/100, scoring 28% on SWE-bench Verified, 43.8% on GPQA Diamond, 67.5% on MMLU-Pro, and 1225 pts on Chatbot Arena Elo.
What is Pixtral 12B best used for?+
Pixtral 12B is best suited for: documento OCR, visual inspección agentes, y abierto multimodal canalizaciones. Assessment: de Mistral breakthrough entry en de pesos abiertos multimodal computer visión.
Mistral AI Family Lineage, Routing Tiers & Upgrade Path
Step down to a lighter sibling for high-volume routing or step up to Mistral AI's flagship reasoning tier:
Mistral Medium 3.5
Mistral Small 4
Devstral 2
Devstral Medium
Devstral Small
Mistral Small 3.2
AI Glossary & Architecture Concepts
View full glossary →Key machine learning concepts and benchmarks relevant to Pixtral 12B:
La unidad básica de texto que procesa un modelo de lenguaje, que normalmente equivale a 3 o 4 caracteres o aproximadamente 0,75 palabras. Por ejemplo, una frase típica se divide en varios tokens.
Un conjunto estandarizado de pruebas para evaluar y comparar objetivamente las capacidades de distintos modelos de IA en tareas como programación (SWE-bench) o matemáticas avanzadas (AIME).
La cantidad máxima de tokens que un modelo puede considerar a la vez, incluyendo tanto la entrada que lee como la salida que genera. Una ventana de 1.000.000 de tokens puede procesar unas 750.000 palabras en una sola petición.
La capacidad de un modelo de detectar la necesidad de una función externa y emitir argumentos JSON estructurados para ejecutar API, búsquedas web o consultas a bases de datos.
Un protocolo abierto desarrollado por Anthropic que permite a los agentes de IA conectarse de forma estandarizada a sistemas de archivos locales, entornos de desarrollo y servicios externos.
Una optimización donde los proveedores almacenan estados de atención previos para prefijos de prompts frecuentes, reduciendo el coste de entrada hasta en un 90% y acelerando la respuesta.
Las puntuaciones de benchmarks, precios por token y límites de tasa evolucionan rápidamente. Verifique siempre las especificaciones actuales en la documentación oficial del proveedor.