Pixtral Large
Mistral's 124B Multimodal Frontier
Mistral's frontier-tier multimodal model — Mistral Large 2 with a dedicated 1B vision encoder. At launch (Nov 2024), it led MathVista (visual math reasoning) and matched closed-source frontier on most visual tasks. Open-weight under the Mistral Research License.
Intelligence
Below avg
Speed
Slow
56 tok/s output
Cost
High
$2.00 in / $6.00 out
Context
128K
Up to 128,000 tokens
How are Intelligence, Speed & Cost bucketed?
Intelligence and Speed buckets are percentile ranks on
Artificial Analysis. Cost buckets are fixed dollar
thresholds keyed off output-token price ($/M out).
Intelligence
- Top 1%≤ 1%
- Top 5%≤ 5%
- Top 10%≤ 10%
- Good≤ 25%
- Medium≤ 50%
- Below avg> 50%
Speed
- Top 1%≥ 345 tok/s
- Top 5%≥ 237 tok/s
- Top 10%≥ 196 tok/s
- Good≥ 146 tok/s
- Medium≥ 90 tok/s
- Slow< 90 tok/s
Cost
- Freeopen weights · self-host
- Low< $1 / M out
- Moderate$1–5 / M out
- High≥ $5 / M out
Why it matters
Showed that visual reasoning (chart understanding, diagram parsing, document QA) could beat closed frontier models on specific tasks while shipping open weights — particularly for math-heavy visual benchmarks.
Core Capabilities
Long Documents
Handles entire codebases, books, and multi-doc RAG.
Multimodal
Combines text, vision, and audio in one model.
Generative
Produces images, video, audio, or other media.
Agent Workflows
Built for tool use and autonomous tasks.
Context Window
128k tokens
≈ 98 pages
4k Chat 聊天
32k Long docs 长文档
128k This model 本模型
400k Multi-doc 多文档
1M Codebase 整个代码库
10M
Availability
API
Not available
Product / App
Not available
Open Source
Released
Enterprise
—
Pricing Model
Free / self-host
Open weights — pay only for compute.
Self-host Capability / Performance
Where this model sits relative to the middle 60% of models in the tree. All scores are 0–10 (higher is better).
Lower 20% Upper 80% This model
Context / memory
Context window size · log-scaled
6.0
9.0
6.0
Lower 20% 20th percentile — 20% of models score below this This model Where the current model lands Upper 80% 80th percentile — only 20% of models score above this
Percentile boundaries are computed across every model in the tree that reports the underlying benchmark for each capability.
What it feels like
- 124B open-weights multimodal model built on Mistral Large 2 — frontier-level image understanding
- Best open-weights model on the LMSys Vision Leaderboard at release — beats nearest competitor by ~50 ELO
- Pixtral 12B (precursor) hit 52.5% on MMMU at 12B scale, beating much larger models
- Charts, figures, document QA, and instruction following are the noted strengths
- Inherits Mistral Large 2's text capabilities — multimodal without compromising text
- Inference 30-50% slower than text-only models at the same parameter count — vision encoder overhead
Best use cases
- Open-weights vision tasks where Claude Sonnet / GPT-4o aren't acceptable
- Document QA, chart understanding, and figure analysis at frontier quality
- Self-hosted multimodal RAG and OCR-heavy workflows
- Research and academic work needing an open-weights vision baseline
Tools to try
Not ideal for
- Latency-sensitive vision workflows — encoder overhead is significant
- Edge / single-GPU deployments at 124B scale
- Tasks needing precise counting or fine-grained visual measurement (a noted weakness)