Model vs model
Kimi K2.7 Code vs Mistral Medium 3.5
At a glance
Longer bar means the bigger number.
Open or closed · can you download the model and run it yourself?
Blended price · per 1M tokens, lower is better
Intelligence Index · Artificial Analysis, higher is better
Where they sit among all 45 models
Kimi K2.7 Code ranks 24th of the 41 models with an Intelligence Index, Mistral Medium 3.5 34th. Up and to the left is smarter for less money. Hover or tap a dot for its numbers, then click it to compare that model with Kimi K2.7 Code.
Which one to pick
Pick Kimi K2.7 Code if
- you want the lower bill ($1.71 vs $3 per 1M tokens blended)
- you want the higher overall score (25.8 vs 14.2)
- you care about Humanity's Last Exam (hard reasoning): 35 vs 13.8
- you care about GPQA Diamond (science): 89.6 vs 74.8
Pick Mistral Medium 3.5 if
- you care about IFBench (instruction following): 68.8 vs 63.1
- you need to send PDFs
How smart is each one?
Of the 16 scores both models have, Kimi K2.7 Code is ahead on 15 and Mistral Medium 3.5 on 1.
Every intelligence score we have
Out of 100, higher is better. Kimi K2.7 Code Mistral Medium 3.5
Overall
Intelligence Index
25.814.2
By area
Coding
60.846.9Agentic work
217
By field of work
Finance & accounting
28.914Legal
27.716.1Healthcare & medicine
25.914.8Economics
36.520.4Engineering
29.117.3Strategy & operations
2811.9
Benchmarks
Factual knowledge · SimpleQA
36.5Mistral Medium 3.5: not scored yet
Hard reasoning · Humanity's Last Exam
3513.8Science · GPQA Diamond
89.674.8Competition math · OTIS Mock AIME
95.6Mistral Medium 3.5: not scored yet
Research math · FrontierMath
54Mistral Medium 3.5: not scored yet
Scientific coding · SciCode
47.840.2Terminal tasks · Terminal-Bench 4.0
10Tool use · τ-Bench Banking
20.215.1Instruction following · IFBench
63.168.8Long documents · AA-LCR
79.369.3Uncensored · UGI
49.9Kimi K2.7 Code: not scored yet
Our tests
Next.js coding tasks · Vercel's Next.js evals, % of tasks passed
74Mistral Medium 3.5: not scored yet
The specs
| Spec | Kimi K2.7 Code | Mistral Medium 3.5 |
|---|---|---|
| Input price | $0.95 per 1M | $1.50 per 1M |
| Output price | $4 per 1M | $7.50 per 1M |
| Cost per test task | $0.5412 | $0.5020 |
| Context window | 262K | 262K |
| Max output | 262K | 262K |
| Accepts | Text, Image, Video | Text, Image, PDF |
| Open weights | Yes | Yes |
| Lab | Kimi | Mistral AI |
| Released | Jun 2026 | Apr 2026 |
| Knowledge cutoff | Jan 2025 | Not published |
Blended price weighs 3 parts input to 1 part output. List prices from models.dev. Cost per test task from Artificial Analysis.
Questions
Is Kimi K2.7 Code or Mistral Medium 3.5 cheaper?
Kimi K2.7 Code is cheaper. It charges $0.95 per 1M input tokens and $4 per 1M output tokens, against $1.50 and $7.50 for Mistral Medium 3.5. Blended 3 parts input to 1 part output, that is $1.71 against $3 per 1M tokens.
Are Kimi K2.7 Code and Mistral Medium 3.5 open source?
Both are open-weights models: Kimi and Mistral AI publish the weights, so you can download them and run them on your own hardware, or call them through any host's API. Check each model's license before you ship it.
Which is smarter, Kimi K2.7 Code or Mistral Medium 3.5?
Kimi K2.7 Code scores higher on the Artificial Analysis Intelligence Index, 25.8 against 14.2 for Mistral Medium 3.5. The biggest differences are on Humanity's Last Exam (35 vs 13.8) and GPQA Diamond (89.6 vs 74.8).
Which has the longer context window?
Both accept 262K tokens.
Can I compare Kimi K2.7 Code and Mistral Medium 3.5 on my own prompt?
Yes. ModelCompare sends one prompt to both models at once and shows each answer side by side with what it cost. You can try it free without an account.
Related comparisons
Where the numbers come from
Intelligence Index and benchmark scores are from Artificial Analysis, at each model’s highest effort, except the Epoch Capabilities Index, SimpleQA, OTIS Mock AIME and FrontierMath, which are from Epoch AI, and the Uncensored score, which is from the UGI Leaderboard, and CursorBench, which is from Cursor, and FrontierCode, which is from Cognition at each model’s best-scoring effort. The hallucination test is ours: questions with one exact answer, fake evidence pushed against right answers and requests that tempt a model to name packages that don’t exist. Prices are list prices from models.dev. Example answers come from our own test runs. Updated November 13, 2026.