Intelligence index
60/ 100
vs all models18th pctile
Composite of MMLU, GPQA, MATH & HumanEval
Speed
70tok/s
vs all models32th pctile
Median across providers, steady state
Blended price
$3.00/ 1M tokens
vs all models36th pctile
3:1 input:output blend
At a glance
- Context window
- 128k tokens
- Max output
- 8k tokens
- Input price
- $2.00 / 1M tokens
- Output price
- $6.00 / 1M tokens
- Time to first token
- 0.6s
- Input modalities
- text
- Output modalities
- text
- License
- Proprietary
- Provider
- Mistral
Benchmark scores
Public scores from each provider; bars compare this model against the leader in each benchmark.
MMLU
General knowledge across 57 subjects
84.0
leader: 91.8
MMLU Pro
Harder MMLU successor with more reasoning
69.4
leader: 80.0
GPQA
Graduate-level science Q&A
48.0
leader: 78.0
MATH
Competition mathematics
71.5
leader: 94.8
HumanEval
Python code generation pass@1
88.0
leader: 95.8
Strengths
- EU data residency
- Strong multilingual
- Good code
Weaknesses
- Behind US labs on raw benchmark scores
Best for
- EU compliance
- Multilingual apps
- Code tooling
Models you should also evaluate
Anthropic
Claude 4 Opus
Anthropic’s 2026 flagship — best-in-class on code and long-horizon agents.
81 intel50 tok/s$16.00 /1M
OpenAI
GPT-5.5
OpenAI’s 2026 flagship — strongest at reasoning, coding and tool use.
82 intel95 tok/s$7.50 /1M
Google
Gemini 2.5 Pro
2M-token context + native multimodality — unbeatable for huge docs.
78 intel110 tok/s$2.19 /1M
Mistral Large 2 — frequently asked questions
Mistral Large 2 is a large language model from Mistral, released on 24 July 2024. EU-hosted frontier — multilingual and strong on code.
Need help choosing between models?
Compare every option in one sortable table — intelligence, speed and price on a single page.