Is Medium 3.5 available with Zero Data Retention on Elosia?

Medium 3.5 runs on Elosia's standard backend; ZDR is available on many other models in the catalog.

How much does Medium 3.5 cost on Elosia?

Medium 3.5 is billed by usage at $1.5 per 1M input tokens and $7.5 per 1M output tokens, within your Elosia plan.

What is Medium 3.5's context window?

Medium 3.5 supports a context window of up to 262,144 tokens.

Is Medium 3.5 hosted in the EU?

Yes, Medium 3.5 is available on Elosia's EU-hosted (EdenAI) backend for enterprise organizations.

When should you choose Medium 3.5?

Medium 3.5 is a balanced model: a strong quality/cost trade-off for everyday production work.

Medium 3.5

MistralBalanced

ThinkingTool UseVisionStructured Output

About this model

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...

Performance Tier

Balanced

Medium 3.5 is a balanced model from Mistral : strong performance at a reasonable price.

Strong cost-performance ratio. Reliable for most professional use cases without premium pricing.

Pricing

This model is included in Elosia plans

Moderate

Moderate cost. A balanced choice for regular use.

Type	per 1M tokens
Input (prompt)	$1.50
Output (completion)	$7.50

Capabilities

Context Length262K

Max Output Tokens210K

TokenizerMistral

Inputtext, image, file

Outputtext

Release DateApril 30, 2026

Benchmarks

General Intelligence

MMLU

Not reported

Mathematics

MATH-500

Not reported

AIME 2025

86.3%

Programming

HumanEval

Not reported

SWE-bench Verified

77.6%

Where does Medium 3.5 stand?

Compare its performance index against every other model.

View the performance leaderboard

Recommended Use Cases

CodingAnalysisMathematicsGeneral Chat

Strengths

Strong agentic coding performance: 77.6% on SWE-bench Verified, ahead of Claude Sonnet 4.5 (77.2%) and Devstral 2 (72.2%)
Excellent math reasoning: 86.3% AIME 2025 avg@16, on par with Claude Sonnet 4.5/4.6 (86.7/86.9%)
Solid instruction following: 95.8% on Collie, ahead of Claude Sonnet 4.5 (90.5%)
Strong agentic tool-use across τ³ benchmarks: 91.4% Telecom, 76.1% Retail, 72.0% Airline
256K context window, native multimodal vision, and configurable reasoning mode
Open weights under modified MIT license, EU-based provider for data sovereignty

Limitations

No published scores on MMLU, GPQA Diamond, MATH-500, or HumanEval, limiting comparison on classic general reasoning benchmarks
Web browsing is the weakest axis: 48.6% on BrowseComp, well behind Qwen3.5 (78.6%), GLM-5 (74.9%), and Kimi K2.5 (74.7%)
Verbose outputs (~5× the median token count) inflate cost per query
Premium pricing for an open-weight model: $1.50 / $7.50 per million input/output tokens