AI model comparison
Mistral Medium 3.5 vs Sonar Pro
Side by side comparison of Mistral Medium 3.5 and Sonar Pro on pricing, benchmark scores, context window, and more.
Mistral Medium 3.5 is a dense 128B instruction-following model from Mistral AI.
Mistral Medium 3.5 is scheduled to be retired on January 31, 2027. Consider Mistral Large 4 instead.
- Input / 1M
- $1.50
- Output / 1M
- $7.50
- Context
- 262K
Note: Sonar Pro pricing includes Perplexity search pricing.
- Input / 1M
- $3
- Output / 1M
- $15
- Context
- 200K
Comparing image generators? Compare image models
At a glance
Quick verdict
How the two models stack up on the things people ask about most.
Lower price
Mistral Medium 3.5
2x cheaper for a typical chat
Higher intelligence score
Sonar Pro
7.6 vs 4.6 on Artificial Analysis
Larger context window
Mistral Medium 3.5
262K vs 200K tokens
Newer release
Mistral Medium 3.5
Released April 30, 2026
Benchmarks
Benchmark scores
Independent scores from Artificial Analysis. Higher is better.
Intelligence index
Coding index
Agentic index
Pricing
Mistral Medium 3.5 vs Sonar Pro API pricing
Per-token API rates. Cheaper option highlighted.
| Metric | ||
|---|---|---|
| Input tokensPer 1M tokens | $1.50 | $3 |
| Output tokensPer 1M tokens | $7.50 | $15 |
| Cached input (read)Per 1M tokens | Not available | Not available |
| Cache writePer 1M tokens | Not available | Not available |
| Web searchPer request | Not available | $0.005 |
What it costs in practice
Estimated cost per 1,000 requests at standard rates. Coding and long-document figures also show the cost when the input is already cached.
Chat message
2K in, 500 out
Coding task
30K in, 4K out
Long document summary
150K in, 2K out
Specs
Context window and capabilities
How much each model can read, how much it can write, and what it accepts as input.
| Metric | ||
|---|---|---|
| Context window | 262K tokens | 200K tokens |
| Max output | 210K tokens | 8K tokens |
| Input types | File, Image, Text | Image, Text |
| Output types | Text | Text |
| Reasoning effort levels | High, None | Not available |
| Default reasoning effort | High | Not available |
| Release date | April 30, 2026 | March 7, 2025 |
Features
Supported features
API features available for each model.
| Metric | ||
|---|---|---|
| Tool calling | Supported | Not supported |
| Structured outputs | Supported | Not supported |
| JSON mode | Supported | Not supported |
| Reasoning | Supported | Not supported |
| Temperature | Supported | Supported |
| Stop sequences | Supported | Not supported |
| Deterministic seed | Supported | Not supported |
| Verbosity control | Not supported | Not supported |
Our take
Mistral Medium 3.5 is 2x cheaper for a typical chat. Sonar Pro scores higher on the Artificial Analysis Intelligence Index (7.6 vs 4.6). Mistral Medium 3.5 has the larger context window (262K vs 200K tokens). The right pick depends on your workload, so the quickest way to settle it is to send the same prompt to both and compare.
FAQ
Frequently asked questions
Is Mistral Medium 3.5 or Sonar Pro cheaper?
Mistral Medium 3.5 is cheaper for a typical chat (2K input tokens and 500 output tokens). Mistral Medium 3.5 costs $1.50 per 1M input tokens and $7.50 per 1M output tokens, while Sonar Pro costs $3 per 1M input tokens and $15 per 1M output tokens.
Which has a bigger context window, Mistral Medium 3.5 or Sonar Pro?
Mistral Medium 3.5 supports up to 262K tokens of context, compared with 200K for Sonar Pro.
Is Mistral Medium 3.5 smarter than Sonar Pro?
On the Artificial Analysis Intelligence Index, Mistral Medium 3.5 scores 4.6 and Sonar Pro scores 7.6, putting Sonar Pro ahead. Benchmarks don't capture everything, so the best test is running your own prompts through both.
Can I use Mistral Medium 3.5 and Sonar Pro at the same time?
Yes. Shortcut Chat sends one prompt to multiple models at once, so you can see Mistral Medium 3.5 and Sonar Pro answer side by side and keep the better response.
Why choose? Ask both.
Send one prompt to Mistral Medium 3.5 and Sonar Pro at the same time. Compare the answers side by side and keep the best one.