Requesty
Mistral AI SAS logo
Mistral AI SAS

mistral-small-2503

Mistral Small 3.1 (2503) is a powerful 24-billion parameter multimodal AI model developed by Mistral AI that processes both text and image inputs. Designed for enterprise efficiency and fast-response agents, it is highly sought after for its exceptional long-context processing, function calling, and local deployment capabilities.

Tool callingJSON schema
Try mistral-small-2503CompareDocs

Specifications

Context window33K tokens
Max outputN/A
API typechat
AddedJun 29, 2026
Model IDmistral/mistral-small-2503
Data retentionYes (30 days)
Used for trainingNo
Provider location🇪🇺 EU
Privacy policyMistral Privacy Policy→

Benchmarks

Released 2023-12-11
GPQA Diamondreasoning
34.9%

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

Intelligence Indexreasoning
3.6%

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
$0.10
Output / 1M
$0.30
Cache write
N/A
Cache read
N/A
Estimated cost
100K input + 10K output$0.0130
1M input + 100K output$0.13
10M input + 1M output$1.30

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-small-2503.

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-small-2503",
    messages=[
        {"role": "user", "content": "Explain quantum computing in one paragraph."},
    ],
)

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

Other Mistral AI SAS models

leanstral-1-5
262K ctxFree/1M
codestral-latest
131K ctx$0.30/1M
devstral-latest
256K ctx$0.40/1M
open-mistral-7b
33K ctx$0.25/1M
mistral-medium-latest
131K ctx$0.40/1M
devstral-small-2507
131K ctx$0.10/1M

Frequently asked questions

How much does mistral-small-2503 cost?
mistral-small-2503 is priced at $0.10 per million input tokens and $0.30 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-small-2503?
mistral-small-2503 has a context window of 33K tokens. That's roughly 44 words of input you can fit in a single prompt.
How does mistral-small-2503 perform on benchmarks?
mistral-small-2503 scores 49.1% on MMLU Pro, 34.9% on GPQA Diamond, 11.8% 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-small-2503 do?
mistral-small-2503 supports tool calling, structured outputs (JSON schema). You can call it through any OpenAI-compatible client by pointing base_url to Requesty.
How do I use mistral-small-2503 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-small-2503". The Quickstart above shows Python, JavaScript and cURL snippets.
Can I run mistral-small-2503 through Requesty?
Yes. mistral-small-2503 runs through Requesty's OpenAI-compatible API, served from Mistral AI SAS. You do not host the model yourself: point base_url at Requesty, set the model to "mistral/mistral-small-2503", and requests are routed to the upstream provider with automatic failover. The same key gives you 400+ other models too.

Access mistral-small-2503 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