- Mistral Nemo is a free model from Mistral AI that you can run on your own computer. In our tests it's not one we'd recommend right now: 23 out of 100, #48 of 56.
- It solved 3 of 30 coding jobs and scored 36 on reading documents. On our hardest tasks it scored 13.
- Runs on a 12 GB graphics card or a Mac with 16 GB.
Coding
Our coding test is 30 programming jobs, from small ones like reading time durations or cleaning up messy data to harder ones like a config-file parser or a double-entry ledger. We run each answer against tests the model never sees, and a job only counts if everything passes. Mistral Nemo got 3 of 30 right. The best local coders solved 29 of 30.
Reading documents
The second test hands the model things like an expense claim thread, a pay stub or an insurance statement, and asks for specific numbers and dates. Many questions need a bit of math, or noticing a correction further down the email. Mistral Nemo scored 36; the best model scored 100.
| Test | Score | Public questions | Secret questions |
|---|---|---|---|
| Coding | 10 | 14 | 9 |
| Reading documents | 36 | 51 | 32 |
| Decisions | 78 | 80 | 77 |
On the 18 hardest tasks (included in the scores above) it scored 13. This number separates the top models.
This model did much better on our public questions than on the secret ones. Sometimes that's luck; it can also mean the model has seen the public questions before.
We tested the full-size version online. The smaller versions you download usually score a little lower.
Can your computer run it?
You can download Mistral Nemo in several sizes. Smaller versions fit on more computers and lose a little quality, so pick the biggest one that fits.
| Your computer | Runs it? | Version to download |
|---|---|---|
| Graphics card, 8 GB | No | – |
| Graphics card, 12 GB | Yes | Q5 |
| Graphics card, 16 GB | Yes | Q8 |
| Graphics card, 24 GB | Yes | Q8 |
| Graphics card, 32 GB | Yes | Q8 |
| Mac, 16 GB | Yes | Q4 |
| Mac, 24 GB | Yes | Q8 |
| Mac, 32 GB | Yes | Q8 |
| Mac, 48 GB | Yes | Q8 |
| Mac, 64 GB | Yes | Q8 |
| Mac, 96 GB | Yes | Q8 |
| Mac, 128 GB | Yes | Q8 |
For normal-length chats. Long documents need more memory. Q4 or Q8? · How we worked this out
How to try it
The easiest way is Ollama. Install it, then run:
ollama run mistral-nemo Want the strongest all-rounder instead? Qwen3.6 27B scored 97.
Its actual answers
What Mistral Nemo wrote for our public questions, word for word.
Coding: INI parser (hard) Didn't work
```python
import re
from typing import Dict, Optional
def parse_ini(text: str) -> Dict[str, Dict[str, str]]:
sections: Dict[str, Dict[str, str]] = {}
current_section: Optional[str] = None
default_section: Dict[str, str] = {}
lines = text.splitlines()
i = 0
while i < len(lines):
line = lines[i].strip()
if not line or line.startswith((';', '#')):
i += 1
continue
match = re.match(r'\s*\[([^]]+)\]', line)
if match:
current_section = match.group(1).lower()
if current_section == 'default':
default_section = sections.pop('default', {})
else:
sections[current_section] = default_section.copy()
i += 1
continue
match = re.match(r'\s*([^=:#;]+)\s*(=|:)\s*(.*)', line)
if not match:
raise ValueError(f"Malformed line: {line!r}")
key, sep, value = match.groups()
key = key.lower().strip()
if current_section is None:
raise ValueError(f"No section for key: {key!r}")
if key in sections[current_section]:
i += 1
continue
if value.startswith('"') and value.endswith('"'):
value = value[1:-1]
else:
value = value.strip()
sections[current_section][key] = value
while i + 1 < len(lines) and lines[i + 1].strip() and lines[i + 1].strip()[0].isspace():
value += '\n' + lines[i + 1].strip()
i += 1
i += 1
return sections
``` Coding: TTL LRU (hard) Didn't work
```python
import time
from collections import OrderedDict
class TTLCache:
def __init__(self, capacity: int, ttl: float, clock):
self.capacity = capacity
self.ttl = ttl
self.clock = clock
self.cache = OrderedDict()
self.stats = {"hits": 0, "misses": 0, "evictions": 0, "expirations": 0}
def put(self, key, value):
now = self.clock()
if key in self.cache:
self.cache.move_to_end(key)
self.cache[key] = (now + self.ttl, value)
else:
if len(self.cache) >= self.capacity:
self._evict(now)
self.cache[key] = (now + self.ttl, value)
def get(self, key, default=None):
now = self.clock()
if key in self.cache and self.cache[key][0] > now:
self.cache.move_to_end(key)
return self.cache[key][1]
else:
self.stats["expirations"] += 1
self.stats["misses"] += 1
del self.cache[key]
return default
def __len__(self):
now = self.clock()
return sum(1 for k, v in self.cache.items() if v[0] > now)
def stats(self):
return self.stats
def _evict(self, now):
while self.cache:
key = next(iter(self.cache))
if self.cache[key][0] > now:
break
self.stats["evictions"] += 1
del self.cache[key]
``` Coding: Parse duration Didn't work
```python
import re
def parse_duration(s: str) -> int:
if not s.strip():
raise ValueError("Empty or blank string")
units = {'h': 3600, 'm': 60, 's': 1}
pattern = r'^(\d+[hms]?)*$'
if not re.match(pattern, s, re.IGNORECASE):
raise ValueError("Invalid duration format")
parts = re.split(r'(\d+)', s)
parts = [p for p in parts if p]
for i in range(0, len(parts), 2):
value = int(parts[i])
unit = parts[i + 1].lower()
if unit not in units:
raise ValueError(f"Unknown unit: {unit}")
if i > 0 and units[unit] < units[prev_unit]:
raise ValueError("Units out of order")
total_seconds += value * units[unit]
prev_unit = unit
return total_seconds
``` Coding: Summarize ranges Worked
```python
def summarize_ranges(nums: list[int]) -> str:
nums = sorted(set(nums))
result = []
i = 0
while i < len(nums):
j = i + 1
while j < len(nums) and nums[j] == nums[j - 1] + 1:
j += 1
if j - i >= 3:
result.append(f"{nums[i]}..{nums[j - 1]}")
elif j - i == 2:
result.append(f"{nums[i]},{nums[j - 1]}")
else:
result.append(str(nums[i]))
i = j
return ",".join(result)
``` Coding: Top customers Didn't work
```python
from decimal import Decimal, InvalidOperation
from collections import defaultdict
def top_customers(rows: list[dict], n: int) -> list[tuple[str, float]]:
customers = defaultdict(Decimal)
for row in rows:
name = row.get("customer", "").strip()
if not name:
continue
try:
amount = Decimal(row["amount"].replace(",", ""))
except InvalidOperation:
continue
customers[name.lower()] += amount
return sorted(customers.items(), key=lambda x: (-x[1], x[0]))[:n]
``` Coding: Slugify Didn't work
```python
import re
import unicodedata
def slugify(title: str, max_len: int = 50) -> str:
title = unicodedata.normalize('NFKD', title).encode('ascii', 'ignore').decode('utf-8')
title = re.sub(r'[^a-z0-9]+', '-', title.lower())
title = re.sub(r'-+', '-', title).strip('-')
words = title.split('-')
if len(title) > max_len:
title = '-'.join(words[:max_len // len(words[0])]) if max_len % len(words[0]) == 0 else '-'.join(words[:max_len // len(words[0]) + 1])
title = title[:max_len]
return title.rstrip('-')
``` Coding: Token bucket Didn't work
```python
import time
class TokenBucket:
def __init__(self, capacity: float, refill_per_sec: float, clock: callable):
self.capacity = capacity
self.refill_per_sec = refill_per_sec
self.clock = clock
self.tokens = capacity
self.last_refill = self.clock()
def allow(self, cost: float = 1) -> bool:
self.refill()
if self.tokens >= cost:
self.tokens -= cost
return True
return False
@property
def tokens(self) -> float:
self.refill()
return self._tokens
def refill(self) -> None:
elapsed = self.clock() - self.last_refill
self._tokens = min(self.capacity, self._tokens + elapsed * self.refill_per_sec)
self.last_refill = self.clock()
``` Decisions: Refund window (hard) 50% right
{
"outcome": {
"full_refund": 0.0,
"partial_refund": 0.8,
"store_credit": 0.2,
"nothing": 0.0
},
"defective": {
"true": 0.0,
"false": 1.0
}
} Decisions: Incident matrix (hard) 0% right
{
"severity": {
"0": 0.0,
"1": 0.4,
"2": 0.5,
"3": 0.1
},
"page": {
"true": 0.9,
"false": 0.1
}
} Decisions: Tool followup (hard) 100% right
{
"tool": {
"calendar": 1.0
},
"confirm": {
"true": 1.0
}
} Decisions: Legit security alert (hard) 100% right
{
"phishing": {
"true": 0.0,
"false": 1.0
},
"action_needed": {
"true": 0.0,
"false": 1.0
}
} Decisions: Meeting slot (hard) 50% right
{
"slot": {
"A": 0.0,
"B": 0.0,
"C": 1.0,
"D": 0.0
},
"raj_last": {
"true": 1.0,
"false": 0.0
}
} Decisions: Review mixed (hard) 100% right
{
"hardware": {
"true": 1.0
},
"support": {
"true": 1.0
}
} Decisions: Support checkout down 67% right
{
"department": {
"billing": 0.8,
"technical": 0.2,
"account": 0,
"shipping": 0,
"sales": 0
},
"urgency": {
"0": 0,
"1": 0,
"2": 0.2,
"3": 0.8
},
"outage": {
"true": 1,
"false": 0
}
} Decisions: Refund wrong plan 100% right
{
"department": {
"billing": 0.9,
"account": 0.1,
"technical": 0,
"shipping": 0,
"sales": 0
},
"refund": {
"true": 1,
"false": 0
},
"tone": {
"frustrated": 0.3,
"calm": 0.7
}
} Decisions: Moderation doxxing 100% right
{
"policy": {
"none": 0.2,
"harassment": 0.8
},
"personal_info": {
"true": 1,
"false": 0
}
} Decisions: Route calendar 100% right
{
"tool": {
"calendar": 1.0
},
"confirm": {
"true": 1.0
}
} Decisions: Doc invoice missing due 100% right
{
"doc_type": {
"invoice": 1.0,
"resume": 0.0,
"contract": 0.0,
"bank_statement": 0.0,
"other": 0.0
},
"missing_due_date": {
"true": 1.0,
"false": 0.0
}
} Decisions: Phishing paypal 50% right
{
"phishing": {
"true": 0.95,
"false": 0.05
},
"risk": {
"0": 0.05,
"1": 0.2,
"2": 0.65,
"3": 0.1
}
} Decisions: Pii ssn email 100% right
{
"data_kind": {
"none": 0.0,
"contact": 0.0,
"financial": 0.0,
"government_id": 1.0,
"health": 0.0
},
"sensitive": {
"true": 1.0,
"false": 0.0
}
} Decisions: Review mixed 100% right
{
"sentiment": {
"positive": 0,
"neutral": 0.2,
"negative": 0.8
},
"defect": {
"true": 1,
"false": 0
},
"recommend": {
"true": 0,
"false": 1
}
} Documents: Saas escalator (hard) 20% right
{
"year2_price_per_seat_month": 47.29,
"year3_price_per_seat_month": 45.00,
"year1_invoice": 6600.00,
"year2_invoice": 7560.00,
"addon_months_billed": 10,
"addon_invoice": 7680.00,
"year3_invoice": 8100.00,
"year3_discount_percent": 15.00,
"total_contract_value": 22340.00,
"contract_end_date": "2027-02-28"
} Documents: Expense thread 75% right
{
"employee_id": "EMP-20417",
"destination_city": "Lisbon",
"trip_start": "2025-02-24",
"trip_end": "2025-02-27",
"approved_items": [
{
"date": "2025-02-24",
"category": "airfare",
"amount_usd": 1184.6
},
{
"date": "2025-02-24",
"category": "ground_transport",
"amount_usd": 36.0
},
{
"date": "2025-02-25",
"category": "meals",
"amount_usd": 230.08
},
{
"date": "2025-02-26",
"category": "lodging",
"amount_usd": 466.24
},
{
"date": "2025-02-27",
"category": "ground_transport",
"amount_usd": 44.28
}
],
"rejected_item_count": 2,
"per_diem_days": 3,
"per_diem_usd": 195.0,
"total_reimbursable_usd": 2135.12,
"approver_email": "priya.raman@corvane.com"
} Documents: Lease amendment 67% right
{
"tenants": ["Marcus Lin", "Sofia Lin"],
"landlord": "Ridgeline Property Group LLC",
"zip": "97205",
"lease_end": "2025-11-30",
"original_monthly_rent": 2150,
"monthly_rent_from_2025_06_01": 2242,
"late_fee_from_2025_06_01": 112.1,
"security_deposit": 2150,
"total_pet_deposits": 800,
"total_monthly_payment_july_2025": 2327,
"move_in_payment": 3350
} Documents: Ticket SLA 82% right
{
"ticket_id": "48213",
"account_id": "ACC-7731",
"open_issue": "inventory_sync",
"resolved_issues": ["billing_address"],
"affected_orders": ["SO-99812", "SO-99820", "SO-99827"],
"priority": "P2",
"sla_due_local": "2025-09-12T17:30",
"sla_due_utc": "2025-09-12T22:30:00Z",
"reissued_invoice": "INV-2025-0812"
} Documents: Sales footnotes 10% right
{
"q3_total_usd": 15356,
"q2_total_usd": 15464,
"q2_central_originally_reported_usd": 3047,
"q2_to_q3_change_pct": 5.9,
"top_region_q3": "West",
"fastest_growing_region_q1_to_q3": "International",
"regions_declining_q2_to_q3": [],
"international_q3_organic_usd": 1731,
"west_excluding_mountain_q3_usd": 4291
} Size: 12B parameters. First tested OCT 10.
Comments
Sign in with GitHub to comment. Spam and abuse are hidden automatically.