Mistral Medium 3.5
Mistral Medium 3.5 is a model from Mistral AI, listed under Language models. It was released in April 2026. It accepts up to 256k tokens per request.
- Headline price
- No data
- Context window
- 256k tokens
Current prices
No public API price
This model is distributed as open weights. Running cost depends on your own hardware.
Price history
No price change recorded for this model yet.
How it compares
Show data
| Model | Value |
|---|---|
| GPT-6 Astra | 1.05M |
| GPT-6.1 Sol | 1.05M |
| GPT-6 Luna | 1.05M |
| GPT-6 Sol | 1.05M |
| GPT-5.6 Sol | 1.05M |
| GPT-5.6 Terra | 1.05M |
| GPT-5.6 Luna | 1.05M |
| GPT-6 Astra Pro | 1.05M |
| GPT-6 Luna Pro | 1.05M |
| GPT-6 Sol Pro | 1.05M |
| GPT-6.1 Sol Pro | 1.05M |
| GPT-6.1 Terra | 1.05M |
| GPT-6.1 Terra Pro | 1.05M |
| GPT-5.4 | 1.05M |
| GPT-5.4 Pro | 1.05M |
| GPT-5.5 | 1.05M |
| GPT-5.5 Pro | 1.05M |
| Gemini 4 Argon | 1.05M |
| Gemini 3.8 Flash | 1.05M |
| Gemini Omni Flash | 1.05M |
| Muse Wonder | 1.05M |
| Gemini 3.5 Flash-Lite | 1.05M |
| Gemini 3.7 Flash | 1.05M |
| DeepSeek V4 Flash 0731 | 1.05M |
| Muse Spark 1.3 | 1.05M |
| Kimi K3.1 Pro | 1.05M |
| Qwen3.8 2.4T A95B | 1.05M |
| Gemini 3.1 Pro | 1.05M |
| DeepSeek V4 Flash Vision Exp | 1.05M |
| Gemini 3.6 Flash | 1.05M |
| Gemini 3.1 Flash-Lite | 1.05M |
| Gemini 3 Flash | 1.05M |
| Gemini 2.5 Pro | 1.05M |
| Gemini 2.5 Flash | 1.05M |
| Gemini 2.5 Flash-Lite | 1.05M |
| Muse Spark 1.1 | 1.05M |
| Muse Spark 1.2 | 1.05M |
| GPT-4.1 | 1.05M |
| GPT-4.1 Mini | 1.05M |
| GPT-4.1 nano | 1.05M |
| Claude Fable 5.1 | 1M |
| Claude Opus 5.5 | 1M |
| Claude Sonnet 5.5 | 1M |
| Claude Haiku 5.5 | 1M |
| Grok 4.3 | 1M |
| DeepSeek V4 Pro | 1M |
| DeepSeek V4.1 Flash | 1M |
| Kimi K3 | 1M |
| Claude Mythos 5.1 | 1M |
| Grok 4.20 Reasoning | 1M |
| Grok 4.20 Non-Reasoning | 1M |
| Grok 4.20 Multi-Agent | 1M |
| Z.ai GLM 5.3 (on Mistral) | 1M |
| Qwen3.8 27B | 1M |
| Qwen3.8 Max Prime | 1M |
| Mistral Large 4 | 1M |
| Claude Fable 5 | 1M |
| Claude Mythos 5 | 1M |
| Claude Opus 5 | 1M |
| Claude Sonnet 5 | 1M |
| Claude Opus 4.8 | 1M |
| Claude Opus 4.7 | 1M |
| Claude Opus 4.6 | 1M |
| Claude Sonnet 4.6 | 1M |
| Grok 4.7 | 500k |
| Grok 4.6 | 500k |
| Grok 4.5 | 500k |
| GPT-5.6 Cyber | 400k |
| Chat Latest | 400k |
| GPT-5 | 400k |
| GPT-5 Mini | 400k |
| GPT-5 nano | 400k |
| GPT-5 Pro | 400k |
| GPT-5.2 | 400k |
| GPT-5.2 Pro | 400k |
| GPT-5.4 Mini | 400k |
| GPT-5-Codex | 400k |
| GPT-5.1 | 400k |
| GPT-5.1-Codex | 400k |
| GPT-5.1-Codex Mini | 400k |
| GPT-5.1-Codex-Max | 400k |
| GPT-5.2-Codex | 400k |
| GPT-5.3-Codex | 400k |
| GPT-5.4 nano | 400k |
| Grok Build 0.1 | 256k |
| Mistral Small 4 | 256k |
| Ministral 3 14B | 256k |
| Ministral 3 8B | 256k |
| Ministral 3 3B | 256k |
| Command A | 256k |
| Command A Reasoning | 256k |
| North Mini Code | 256k |
| Kimi K2.7 Code | 256k |
| Mistral Large 3 | 256k |
| Kimi K2.6 | 256k |
| Claude Opus 4.5 | 200k |
| Claude Haiku 4.5 | 200k |
| o3 | 200k |
| o3-pro | 200k |
| Claude Sonnet 4.5 | 200k |
| codex-mini-latest | 200k |
| o1 | 200k |
| o1-pro | 200k |
| o3-deep-research | 200k |
| o3-mini | 200k |
| o4-mini | 200k |
| o4-mini-deep-research | 200k |
| gpt-oss-120b | 131k |
| gpt-oss-20b | 131k |
| Command A+ | 128k |
| Codestral | 128k |
| Aya Expanse 32B | 128k |
| Command A Vision | 128k |
| Command R7B | 128k |
| Command R (08-2024) | 128k |
| Command R+ (08-2024) | 128k |
| ChatGPT-4o | 128k |
| GPT-4 Turbo | 128k |
| GPT-4 Turbo Preview | 128k |
| GPT-4.5 Preview | 128k |
| GPT-4o | 128k |
| GPT-4o Mini | 128k |
| GPT-4o Mini Search Preview | 128k |
| GPT-4o Search Preview | 128k |
| GPT-5 Chat | 128k |
| GPT-5.1 Chat | 128k |
| GPT-5.2 Chat | 128k |
| GPT-5.3 Chat | 128k |
| o1 Preview | 128k |
| o1-mini | 128k |
| Command R | 128k |
| Command R+ | 128k |
| Voxtral Small | 32k |
| GPT-3.5 Turbo | 16k |
| babbage-002 | 16k |
| davinci-002 | 16k |
| Aya Vision 32B | 16k |
| computer-use-preview | 8k |
| GPT-4 | 8k |
| Tiny Aya Earth | 8k |
| Tiny Aya Global | 8k |
| Mistral Medium 3.5 | 256k |
Benchmarks
| Benchmarks | Score |
|---|---|
| Intelligence Index Artificial Analysis's composite score across about ten independent evaluations (reasoning, knowledge, coding, agentic work, long context), run the same way for every model. The strongest models score in the 50s. | 14.2 |
| Terminal-Bench Hard Hard command-line tasks (building software, configuring systems, fixing environments) completed autonomously in a real terminal. A direct test of agentic coding. | 33.3% |
| SciCode Writing code to solve research-level problems in physics, chemistry, biology and maths, checked by tests. | 40.2% |
| GPQA Diamond Graduate-level questions in biology, physics and chemistry, written to be resistant to web search. PhD holders in the relevant field score around 65%, so scores above that indicate genuine domain reasoning rather than recall. | 74.8% |
| Humanity's Last Exam About 2,500 expert-written questions at the frontier of human knowledge across maths, science and the humanities. Even the best models answer well under two thirds. | 13.8% |
| IFBench (instruction following) Precise, verifiable formatting and content constraints the model has not seen in training. Measures how faithfully it follows instructions. | 68.8% |
| τ²-Bench (agentic tool use) Simulated customer-support conversations where the model must use tools and follow policy while a simulated user acts on their side. Measures reliable multi-turn tool use. | 94.2% |
| MMMU-Pro (visual reasoning) College-level questions that require reading charts, diagrams and figures, with distractors that defeat text-only guessing. | 64.9% |
| Long-context reasoning (AA-LCR) Questions that require reasoning across several long documents at once (around 100k tokens), not just finding one passage. | 69.3% |
Independent evaluations by Artificial Analysis, run the same way for every model, at the reasoning setting shown. Percentages unless marked as an index. Source
Alternatives
Same category, closest price, other providers first.
More from Mistral AI
| # | Model | Input | Output | Context | Released | Capabilities | Price change |
|---|---|---|---|---|---|---|---|
| 1 |
Mistral Small 4
Mistral AI
|
No data | No data | 256k | March 2026 | VisionReasoningTool callingSelf-hostable | |
| 2 |
Ministral 3 14B
Mistral AI
|
No data | No data | 256k | December 2025 | VisionTool callingSelf-hostable | |
| 3 |
Ministral 3 8B
Mistral AI
|
No data | No data | 256k | December 2025 | VisionTool callingSelf-hostable | |
| 4 |
Ministral 3 3B
Mistral AI
|
No data | No data | 256k | December 2025 | VisionTool callingSelf-hostable | |
| 5 |
Z.ai GLM 5.3 (on Mistral)
Mistral AI
|
No data | No data | 1M | Aug 18, 2026 | Tool calling | |
| 6 |
Codestral
Mistral AI
|
No data | No data | 128k | August 2025 | Tool calling |
Spotted something wrong? Report it
Frequently asked questions
What is the context window of Mistral Medium 3.5?
Mistral Medium 3.5 accepts up to 256k tokens per request.
Is Mistral Medium 3.5 still available?
Yes. Mistral Medium 3.5 is generally available through the Mistral AI API.
When was Mistral Medium 3.5 released?
Mistral Medium 3.5 was released in April 2026.
What is the API model ID for Mistral Medium 3.5?
Use mistral-medium-3-5 when calling the Mistral AI API.
What are the alternatives to Mistral Medium 3.5?
Models in the same category at a similar price include Grok 4.3 (xAI), Command A+ (Cohere), Grok Build 0.1 (xAI) and Grok 4.20 Reasoning (xAI).