Chat with Pixtral Large 2502 for free to write code, read charts and answer hard questions. It is Mistral's large multimodal model from February 2025, with 124B parameters for text and images.

Pixtral Large 2502 is a large multimodal chat model made by Mistral AI, released in February 2025 as an update to Pixtral Large 2411. It takes text and images as input and writes text back, so you can chat about documents, charts and photos in the same conversation. Mistral built it as the second model in its multimodal family. The goal is frontier-level image understanding while keeping the strong text skills of Mistral Large 2.
It is an older model from early 2025. Mistral now marks Pixtral Large as deprecated, and its own API version was retired on May 31, 2026, so the 25.02 version is served through Amazon Bedrock.
The model has 124B parameters and a 128,000-token context window, with a maximum output of 102,400 tokens. That context fits at least 30 high-resolution images in one chat. It understands dozens of languages, including English, French and German. It also knows Spanish, Chinese and Japanese, and it was trained on more than 80 programming languages such as Python, Java and C++.
Mistral has published no changelog for 25.02 versus 24.11. AWS's launch notes highlight built-in function calling, JSON output and strong system-prompt adherence for RAG. In addition, Amazon Bedrock offers it as a fully managed serverless model.
AIFreeForever keeps Pixtral Large 2502 free for everyone through ads, partnerships, sponsorships and donations. We rely on partnerships, sponsorships and donations to keep the models free. If you use Pixtral Large 2502 a lot, you can sponsor its free use with a donation through Buy Me a Coffee.
Your first 3 messages to Pixtral Large 2502 are free without sign up. After that, you verify your email with a one-time code to get more messages each day.
Pixtral Large 2502 is sometimes not available on AIFreeForever, and each case has a simple fix.
Pixtral Large 2502 can be at capacity because of high demand, or the day's messages may be used up. When that happens, the chat answers with another model that is available. The chat tells you which model replied, so you always know who wrote the answer.
Your message and the whole conversation stay in place, so you can carry on without starting over. As soon as Pixtral Large 2502 is back, a notice in the chat lets you know, and your next messages go to it again.
Pixtral Large 2502 on AIFreeForever comes with web research, file and picture reading, and website building built into the chat.
When a question needs current facts, the chat searches the web first and shows the sources it used.
Attach a PDF, a Word file or a picture and ask Pixtral Large 2502 about it.
Builds a website from your description, which you can preview, change and download.
Makes an Excel spreadsheet with your columns, formulas and charts.
Writes a resume from your details, ready to download.
Send the same question to other AI models on AIFreeForever and compare what they say.
People use Pixtral Large 2502 to read images and turn them into answers, from receipt maths to chart analysis and structured data.
Plain chat
Plain chat
Plain chat
Plain chat
Plain chat
Plain chat 
Pixtral Large 2502 is an advanced multimodal model that excels at understanding documents, analyzing complex charts, and extracting data from images.
Pixtral Large 2502 is developed by Mistral AI, and there are several official ways to use this model. The first method is Mistral's own chat application, Le Chat. Le Chat previously featured image and document understanding at its 2024 launch. However, Mistral now lists the model as deprecated and has replaced it with newer vision models.
Developers can access the model through the Mistral API using a paid key. Another option is Amazon Bedrock, which runs the model serverless with pay per use pricing. Developers can also reach it through OpenRouter, which routes requests to Amazon Bedrock.
If you want to chat with Pixtral Large 2502 without using a paid account, AIFreeForever provides the model in a simple chat interface. Your first 3 messages to Pixtral Large 2502 are free without sign up. After those initial messages, you verify your email to receive more messages each day.
Pixtral Large 2502 is made by Mistral AI, and its API costs $2.00 per million input tokens and $6.00 per million output tokens. Those rates come from the OpenRouter endpoint listing, so developers building with the model can budget for them before writing any code.
If you want to buy the API yourself, OpenRouter lists providers that serve Pixtral Large 2502, including Amazon Bedrock. The table below compares those providers on price and context length, so you can pick the one that fits your project. On AIFreeForever, you can test Pixtral Large 2502 for free in the chat before you spend anything on the API.
These prices were checked on November 20, 2026 from OpenRouter and OpenRouter.
Pixtral Large 2502 changed little in published features from Pixtral Large 2411. Mistral has published no changelog for the update. The price stayed at $2 input and $6 output per 1M tokens. AWS's launch notes for 25.02 stress built-in function calling, JSON output, and strong system-prompt adherence for RAG. Mistral deprecated the older 2411 on February 27, 2026 and retired it on May 31, 2026, per its model docs. The table below compares Pixtral Large 2502 with the previous version and the models Mistral now lists around it. Mistral Medium 3.5 is the named replacement at $1.5 input and $7.5 output. Mistral Large 4 offers 1M context and reasoning at about a third of the price. Mistral Small 4 combines reasoning, multimodal understanding, and coding at $0.15 input and $0.6 output.
This comparison was extracted from Mistral's docs and OpenRouter.
Opinions about Pixtral Large 2502 are mostly warm on its vision skills, with a few fair questions about its limits and its age.
At launch it was praised as the best open-weights vision model on the LMSys Vision Leaderboard. This set high expectations for the model you can chat with on AIFreeForever.
AWS's authors were impressed that it read French maths diagrams and notation, and that it kept context like a tutor. This is a useful sign for students who upload study material.
InfoQ noted at launch that OCR was a weak spot which Mistral said it would improve later, so scanned text may come back imperfect.
Mistral itself now calls it deprecated and replaced by more powerful vision models, which matters if you want the newest picture understanding available.

Pixtral Large 2502 is an older model, and its limits show next to what current models offer.
Mistral AI has deprecated the model, so it no longer receives updates or maintenance. Expect its answers to reflect an older snapshot of the world.
The model handles 128K tokens of context, far less than the 1M-token models now common. Long books or large codebases will not fit in one chat.
Input covers text and images, with no audio, video or file support and no reasoning mode.
Through OpenRouter the model lacks tool calling and enforced JSON output, which limits automation work.
Mistral built it for general vision tasks, and InfoQ notes it was not designed as a dedicated OCR model.
On OpenRouter, a single provider, Amazon Bedrock, serves it, so capacity can be tight at peak times.
Ask Pixtral Large 2502 a question, paste your code or add a file and get instant answers. Not satisfied? Compare your answers to other models for free.
Chat with Pixtral Large 2502If you want to compare Pixtral Large 2502 with other multimodal models, four stand out. Mistral Small 4 is Mistral's newer model that took over Pixtral's multimodal understanding. GLM 4.6V reads images, charts and documents. Llama 4 Scout is an open multimodal model of the same era, and Ministral 3 8B is a small, current Mistral model for people who want Mistral at low cost.
Yes, you can chat with Pixtral Large 2502 for free on AIFreeForever. Your first 3 messages are free without sign up. After that, you verify your email to get more messages each day. The official API price is $2.00 for input and $6.00 for output per 1M tokens.
Yes, you can access Pixtral Large 2502 without signing up for your initial messages. The first 3 messages are available immediately. To continue using the model, you must verify your email to unlock more messages each day.
Pixtral Large 2502 has a large context window of 128,000 tokens. It also supports a maximum output of 102,400 tokens. This large capacity allows you to process complex and extensive conversations.
Mistral AI created Pixtral Large 2502. The model was released in February 2025. It became available on Amazon Bedrock from April 8, 2025.
Yes, Pixtral Large 2502 can handle both text and image inputs. This capability allows you to provide visual context along with your written prompts.
If Pixtral Large 2502 is currently at capacity, AIFreeForever will offer an alternative model. The chat will continue using the available model, and you will be notified when Pixtral Large 2502 is back online.
AIFreeForever is not affiliated with or endorsed by the model providers. All trademarks and logos belong to their owners.
Tell us if this AI content is offensive, harmful or wrong. We review every report and use it to improve our filters.