Refined 24B open-weights model upgraded with native vision capabilities, 128K context window, and enhanced coding precision.
Technical Specifications
- Provider
- Mistral AI
- Category
- open-source
- Release Date
- 2025-06
- 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)
- mistral-small-3-2-latest
- Supported SDK & API Protocol
- OpenAI-Compatible REST / vLLM / SGLang / HuggingFace TGI
- Lifecycle & Deprecation Status
- Active GA (Guaranteed support through 2027+)
- 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 ($0.1/1M 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
- 15 GB VRAM (4-bit Quantized) · 52 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
Mistral Small 3.2 Cost, Context, Hardware & API Integration
Pay-as-you-go API is cheaper for this volume (saves $11.00/mo vs $20 flat plan).
Break-even point: ~1,111 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 Mistral Small 3.2:
Benchmark Scores & 6-Axis Capability Radar
- SWE-bench Verified
- 34%
- GPQA Diamond
- 48.5%
- MMLU-Pro
- 72.5%
- HumanEval
- 82%
- Terminal-Bench
- 29%
- AIME
- 55%
- Long-Context Reasoning
- 68%
- Chatbot Arena Elo
- 1255pts
Strengths
- Expanded 128K context window with native multimodal image understanding
- Fast 100 tok/s execution speed on commodity hardware
- Permissive open weights for on-premise deployments
Limitations & Trade-offs
- Image resolution capped compared to large vision models
- Lacks deep autonomous chain-of-thought
Editorial Assessment
An impressively versatile 24B multimodal upgrade for cost-conscious architectures.
Best for: High-volume multimodal triage, edge image-text extraction, and responsive chat assistants
Compare Mistral Small 3.2
Frequently Asked Questions about Mistral Small 3.2
How much does the Mistral Small 3.2 API cost per 1M tokens?+
Mistral Small 3.2 by Mistral AI is priced at $0.1 per 1M input tokens and $0.3 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 Mistral Small 3.2 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 Mistral Small 3.2's context window?+
Mistral Small 3.2 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 Mistral Small 3.2 locally on my own GPU (Open Weights)?+
Yes, Mistral Small 3.2 (24B) 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 Mistral Small 3.2 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 Mistral Small 3.2 perform on coding and reasoning benchmarks?+
Mistral Small 3.2 achieves a composite Intelligence Index of 43/100, scoring 34% on SWE-bench Verified, 48.5% on GPQA Diamond, 72.5% on MMLU-Pro, and 1255 pts on Chatbot Arena Elo.
What is Mistral Small 3.2 best used for?+
Mistral Small 3.2 is best suited for: High-volume multimodal triage, edge image-text extraction, and responsive chat assistants. Assessment: An impressively versatile 24B multimodal upgrade for cost-conscious architectures.
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 Medium 3
AI Glossary & Architecture Concepts
View full glossary →Key machine learning concepts and benchmarks relevant to Mistral Small 3.2:
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.
A cost and latency optimization where inference providers store pre-computed attention states for repeated prompt prefixes. Reusing cached tokens typically slashes API input costs by 50% to 90% and cuts TTFT substantially.
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.
Sources: Mistral AI Open Models (June 2025)↗