Claude Sonnet 4.5 vs Mistral Medium 3.5
Benchmarks, API pricing and specs, head to head. Data updated 2026-06-10.
Anthropic · Sep 2025
The model that made 30-hour autonomous coding sessions real — 77.2% SWE-bench Verified and a 1M-token context beta at $3/$15.
Mistral AI · Mar 2026
Europe's strongest model line: frontier-adjacent quality at $0.40/$2 with EU data residency and on-prem options.
The verdict
These two models don't yet share verified results on the benchmarks we track, so judge them on specs, pricing and intended use. On price the gap is dramatic: Mistral Medium 3.5 works out roughly 7.5x cheaper per blended million tokens. Claude Sonnet 4.5 also takes 200K of context versus 131K for Mistral Medium 3.5.
Specs & pricing
| Claude Sonnet 4.5 | Mistral Medium 3.5 | |
|---|---|---|
| modhub Index | 65.5 | — |
| Input price / 1M | $3 | $0.4 |
| Output price / 1M | $15 | $2 |
| Context window | 200K | 131K |
| Max output | 64K | — |
| Open weights | no | no |
| Reasoning model | yes | no |
| Multimodal input | text, image | text, image |
| Knowledge cutoff | Jan 2025 | Dec 2025 |
| Released | Sep 2025 | Mar 2026 |
| Example monthly cost* | $52.50 | $7.00 |
* 10M input + 1.5M output tokens per month at list prices, no caching. Green = better value on that row.
Frequently asked questions
- Which is better, Claude Sonnet 4.5 or Mistral Medium 3.5?
- These two models don't yet share verified results on the benchmarks we track, so judge them on specs, pricing and intended use. On price the gap is dramatic: Mistral Medium 3.5 works out roughly 7.5x cheaper per blended million tokens. Claude Sonnet 4.5 also takes 200K of context versus 131K for Mistral Medium 3.5.
- Which is cheaper, Claude Sonnet 4.5 or Mistral Medium 3.5?
- Claude Sonnet 4.5 costs $3/$15 per million input/output tokens, while Mistral Medium 3.5 costs $0.4/$2. For a typical workload of 10M input and 1.5M output tokens per month, that's $52.50 versus $7.00.
- Which model is better for coding, Claude Sonnet 4.5 or Mistral Medium 3.5?
- We don't yet track SWE-bench Verified results for both models; check their individual pages for coding-related scores.