Requesty
Mistral AI SAS logo
Mistral AI SAS

mistral-medium-3-5

Mistral Medium 3.5 is a dense 128B instruction following model from Mistral AI. It supports text and image inputs with text output, and is designed for agentic workflows, coding, and complex multi step reasoning. It is particularly strong at reliable multi tool calling and long horizon tasks, with a 256K context window, configurable reasoning effort per request, and a custom vision encoder that handles variable image sizes and aspect ratios. Self hostable on as few as four GPUs and available under open weights.

VisionReasoningTool callingJSON schema
Try mistral-medium-3-5CompareDocs

Specifications

Context window262K tokens
Max output—
API typechat
AddedMay 20, 2026
Model IDmistral/mistral-medium-3-5
Data retentionYes (30 days)
Used for trainingNo
Provider location🇪🇺 EU
Privacy policyMistral Privacy Policy→

Benchmarks

Released 2026-04-29
Coding Indexcoding
35.4%

Artificial Analysis Coding Index — a composite of coding evaluations including LiveCodeBench, SciCode and Terminal-Bench.

GPQA Diamondreasoning
74.8%

Graduate-level physics, chemistry & biology questions designed to resist Googling.

Intelligence Indexreasoning
39.2%

Artificial Analysis Intelligence Index — a composite of multiple evaluations measuring overall model capability.

Scores are sourced from official model cards, Artificial Analysis, and public leaderboards. Benchmarks measure specific skills and do not capture every aspect of model quality — always test on your own workload.

Pricing

Input / 1M
$1.50
Output / 1M
$7.50
Cache write
—
Cache read
—
Estimated cost
100K input + 10K output$0.22
1M input + 100K output$2.25
10M input + 1M output$22.50

Requesty charges exactly what the upstream provider charges — no markup, no per-request fees. Prompt caching and smart routing can reduce effective cost by 30-80%.

Quickstart

Drop-in compatible with the OpenAI SDK. Change the base URL, swap in your Requesty API key, and set the model to mistral/mistral-medium-3-5.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
from openai import OpenAI

client = OpenAI(
    api_key="YOUR_REQUESTY_API_KEY",
    base_url="https://router.requesty.ai/v1",
)

response = client.chat.completions.create(
    model="mistral/mistral-medium-3-5",
    messages=[
        {"role": "user", "content": "Explain quantum computing in one paragraph."},
    ],
)

print(response.choices[0].message.content)

Other Mistral AI SAS models

mistral-small-latest
256K ctx$0.15/1M
mistral-small-2603
256K ctx$0.15/1M
devstral-latest
256K ctx$0.40/1M
devstral-small-latest
131K ctx$0.10/1M
pixtral-large-latest
131K ctx$2.00/1M
devstral-medium-2507
131K ctx$0.40/1M

Frequently asked questions

How much does mistral-medium-3-5 cost?
mistral-medium-3-5 is priced at $1.50 per million input tokens and $7.50 per million output tokens when accessed via Requesty. Requesty charges exactly what the upstream provider charges — we don't add markup.
What is the context window of mistral-medium-3-5?
mistral-medium-3-5 has a context window of 262K tokens. That's roughly 350 words of input you can fit in a single prompt.
How does mistral-medium-3-5 perform on benchmarks?
mistral-medium-3-5 scores 94.2% on τ²-Bench, 74.8% on GPQA Diamond, 39.6% on SciCode. See the full benchmark chart above for results across MMLU Pro, GPQA Diamond, SWE-Bench Verified, HumanEval, MATH, AIME, MMMU, and LiveBench.
What can mistral-medium-3-5 do?
mistral-medium-3-5 supports vision input, tool calling, extended reasoning, structured outputs (JSON schema). You can call it through any OpenAI-compatible client by pointing base_url to Requesty.
How do I use mistral-medium-3-5 with the OpenAI SDK?
Install the OpenAI SDK, set base_url to "https://router.requesty.ai/v1", set your API key to your Requesty key, and set the model to "mistral/mistral-medium-3-5". The Quickstart above shows Python, JavaScript and cURL snippets.

Access mistral-medium-3-5 through Requesty

One API key, 400+ models, OpenAI-compatible. No markup on provider prices, automatic failover, and smart caching built-in.

Get started freeAll Mistral AI SAS models