Model comparison · accepted catalogue facts 2026-11-22
Mistral Medium 3.5 vs Mistral Small 3
Neither model has both an equal-or-higher measured score and an equal-or-lower estimated price in this pair. Retaining your current model is a valid outcome.
Both models belong to the Mistral, Codestral, Devstral and Voxtral family. Inclusion does not recommend either model. This is a direct pair comparison; global frontier membership is not its verdict.
Balanced effective USD / 1M tokens · lower is better
$0.058/M
In-Depth Analysis
Mistral Medium 3.5 vs Mistral Small 3: Measured Technical Breakdown
Comprehensive comparison across measured human-preference evaluation, effective production token costs, architectural limits, and migration trade-offs.
Measured Capability (LMArena)
Mistral Medium 3.5 leads by 187.1 Elo
Mistral Medium 3.5: 1420.7 ±6.4 · Mistral Small 3: 1233.6 ±5.88 Statistically significant preference separation
Cost Comparison Across Canonical Production Workloads
Rates represent effective cost per 1M tokens under standard token blends. Cached prompt reads are factored according to workload context reuse.
Workload & Ratio
Mistral Medium 3.5
Mistral Small 3
Cost Difference
Advantage
Balanced3 tokens in per 1 out
$3.00/1M tokens
$0.058/1M tokens
$2.94(98% diff)
Mistral Small 3 costs less
SummariseHuge input, tiny output
$1.80/1M tokens
$0.052/1M tokens
$1.75(97% diff)
Mistral Small 3 costs less
ChatLong cached system prompt
$3.90/1M tokens
$0.062/1M tokens
$3.84(98% diff)
Mistral Small 3 costs less
Code genOutput dominates
$5.10/1M tokens
$0.068/1M tokens
$5.03(99% diff)
Mistral Small 3 costs less
AgenticContext replayed each turn
$2.40/1M tokens
$0.055/1M tokens
$2.35(98% diff)
Mistral Small 3 costs less
Architectural & Operational Specifications
Specification
Mistral Medium 3.5
Mistral Small 3
Operational Note
Context window
262,144 tokens
32,768 tokens
Mistral Medium 3.5 has a 8.0× larger context window
Max output generation
209,715 tokens
16,384 tokens
Mistral Medium 3.5 allows longer single-turn completions
Base input price
$1.50 / 1M tokens
$0.050 / 1M tokens
Mistral Small 3 input rate is cheaper
Base output price
$7.50 / 1M tokens
$0.080 / 1M tokens
Mistral Small 3 output rate is cheaper
Cached input price
Not listed / standard rate
Not listed / standard rate
—
Reasoning & thinking tokens
Supported; standard rates
Standard generation
—
Input modalities
text, image, file
text
—
Tool calling / function use
Supported
Not listed
Mistral Medium 3.5 declares function calling / tool support
Open weights & licence
Proprietary (API-only)
Proprietary (API-only)
—
Provider & reseller market
Mistral · 2 provider offers
Mistral · 2 provider offers
—
When to Choose Mistral Medium 3.5
Higher measured general capability (1420.7 Elo on human evaluation).
Larger context window (262,144 tokens) for long-document or repo-level context.
These settings are included in the share link. The fixed-request scenario and private evaluation data below are excluded.
A ↔ B
Neither model has both an equal-or-higher measured score and an equal-or-lower estimated price in this pair. Retaining your current model is a valid outcome.
Estimated blend: 75% input, 25% output; 0% cached input. Base context tier. Prompt length selects the billing tier. An unlisted cache-read rate uses that tier’s full input rate in this blend; missing required input or output rates prevent an estimate. The fixed-request estimate below validates actual context and output limits and requires a published cache-read rate when cache hits are assumed.
What changes between these models
Quality first, then estimated price. LMArena scores are human-preference measurements, not a benchmark of your application. A missing score is unrated.
Mistral Medium 3.5 · Meets the stated model requirements
Mistral Small 3 · Meets the stated model requirements
Material differences and unknown evidence. Deltas compare each model with model A. Unknown facts remain visible.
Provider counts and precision flags describe catalogue offers. They do not establish the precision, availability, region, tool behaviour or limits of your chosen endpoint. Open each model’s provider table before switching.
Rates below are USD per 1M tokens. A threshold selects that whole input/output rate pair.
Input-token threshold
Input
Output
Cached input
Base, up to and including first threshold
$0.05
$0.08
Unknown
Migration preflight · compare each candidate with model A
Keeping Mistral Medium 3.5 is a valid decision. No pairwise comparison certifies a drop-in replacement.
Mistral Medium 3.5 → Mistral Small 3
Known loss or failed requirement
Context window: 262,144 → 32,768 tokens.
Output limit: 209,715 → 16,384 tokens.
Tool use is not supported by the candidate.
Reasoning is not supported by the candidate.
image input is not supported by the candidate.
file input is not supported by the candidate.
Provider endpoint, serving precision, availability and task success need your own verification. Catalogue model capabilities do not certify an endpoint.
Private to this tab unless you save locally or explicitly export private details. These are estimates, not observed invoices. Cross-model token counts and task success are assumptions.
0.1 extra attempts means 10 extra full attempts per 100 requests. Cache writes, tool charges, taxes and unlisted charges are excluded.
Mistral Medium 3.5Estimated total: $6.75 1,000 full attempts · base rates
Mistral Small 3Estimated total: $0.14 1,000 full attempts · base rates
Prompt and output sensitivity · sampled scenarios
Every point recalculates the entire request shape. Tier-boundary samples include one token below, at and above each threshold. Ordering between samples is not guaranteed. Lowest known cost excludes unknown estimates and is not a replacement recommendation.
Vary input; hold output, requests, cache and retries fixed.
Input tokens
Mistral Medium 3.5
Mistral Small 3
Coverage
2,000
$6.75
$0.14Lowest known
All computed
4,096
$9.894
$0.2448Lowest known
All computed
32,768
$52.902Lowest known
ineligibleInput plus output exceeds the published context window.
Incomplete
131,072
$200.358Lowest known
ineligibleInput plus output exceeds the published context window.
Incomplete
Vary output; hold input, requests, cache and retries fixed.
Output tokens
Mistral Medium 3.5
Mistral Small 3
128
$3.96
$0.11024
2,048
$18.36
$0.26384
8,192
$64.44
$0.75536
Bring your observed evaluation results
JSON is parsed locally. Prompts, filenames, individual responses and extra fields are not retained. Import only aggregate measurements for selected models. Task/version labels do not establish that two evaluations used the same dataset.
Use the exact catalogue ID. Replace the example values with observed measurements. Omit costUsd or latencyMs when they are unknown. No public score is inferred from these observations.
Keep the decision and its evidence
A portable record retains selected facts, complete tier ladders, public requirements and the publication manifest. It also embeds the full permitted public catalogue to verify the input artifact checksum. Files are larger for this reason (up to 2 MB); local storage is limited to 12 records. Team review is local to this browser; there is no hosted collaboration.
Saving is opt-in and includes private review fields. Shared comparison links never include them. Version 1 Check references remain separate and are not overwritten or guessed into this format.
Saved local decisions · 0
No version 2 decision records are stored in this browser.
Questions this page answers
Is Mistral Medium 3.5 a better-and-cheaper replacement for Mistral Small 3?
Neither model has both an equal-or-higher measured score and an equal-or-lower estimated price in this pair. Retaining your current model is a valid outcome.
Which model is higher rated: Mistral Medium 3.5 or Mistral Small 3?
Mistral Medium 3.5 scores higher on LMArena general (1420.7 ±6.4) compared to Mistral Small 3 (1233.6 ±5.9), a margin of 187.1 Elo points. The measured confidence intervals do not overlap, demonstrating a statistically significant preference lead on human evaluations.
Which model is more cost-effective: Mistral Medium 3.5 or Mistral Small 3?
On a standard balanced workload (3 prompt tokens per 1 generated token), Mistral Small 3 costs $0.058 per 1M tokens versus $3.00 for Mistral Medium 3.5. Mistral Medium 3.5 is approximately 52.2× the price of Mistral Small 3. Base input rates are $1.50 for Mistral Medium 3.5 and $0.050 for Mistral Small 3; base output rates are $7.50 for Mistral Medium 3.5 and $0.080 for Mistral Small 3.
How do their context windows and generation limits compare?
Mistral Medium 3.5 provides a 262,144-token context window with a maximum output limit of 209,715 tokens. Mistral Small 3 provides a 32,768-token context window with a maximum output of 16,384 tokens. Mistral Medium 3.5 offers a 8.0× larger context capacity, making it suitable for larger repositories and multi-document workflows.
What capabilities are lost when switching between Mistral Medium 3.5 and Mistral Small 3?
Switching from Mistral Medium 3.5 to Mistral Small 3 involves: Context window: 262,144 → 32,768 tokens; Output limit: 209,715 → 16,384 tokens; Tool use is not supported by the candidate; Reasoning is not supported by the candidate; image input is not supported by the candidate; file input is not supported by the candidate. Run test suites on your specific prompt templates and tool declarations before cutting over production traffic.
Can Mistral Medium 3.5 or Mistral Small 3 be self-hosted or deployed privately?
Both Mistral Medium 3.5 and Mistral Small 3 are proprietary cloud models offered via managed API endpoints. Mistral Medium 3.5 is served by Mistral and 2 catalogue offers, while Mistral Small 3 is served by Mistral and 2 catalogue offers.
What is included in this comparison?
Accepted model capabilities, context and output limits, complete price-tier ladders, independent LMArena measurements and missing evidence. Provider endpoint precision and application success still need verification.
Can I compare a different workload or keep a decision record?
Yes. Change the public workload, prompt length and requirements in the comparison workspace. Fixed-request scenarios, aggregate evaluations and review notes stay local unless explicitly exported.