MMODELYST
Models/Mistral Medium 3.5
MDL

Mistral Medium 3.5

Mistral

Verified against Artificial Analysis · Jul 13, 2026
Capability
81
percentile index
Price /M
$1.5 / $7.5
in / out
Context
262.1K
tokens
Avg score
54.9
7 benchmarks
AA Intelligence
29.9
index
Speed
98 tok/s
TTFT 0.8s
Released
Apr 29, 2026
About

Mistral Medium 3.5 is a open-weights model in the Mistral family from Mistral. Benchmarked on 7 evals, averaging 54.9.

Benchmark scores · 7
avg 54.9 — higher is better
τ²-bench
94.2
GPQA Diamond
74.8
IFBench
68.8
AA Long-Context Reasoning
61
SciCode
39.6
Terminal-Bench Hard
33.3
Humanity's Last Exam
12.8
The research behind this model
all papers mentioning it →

Mentions matched by name in title/abstract — from the arXiv + HF daily corpus.

Details
PublisherMistral
FamilyMistral
Parameters
License
WeightsOpen
ReleasedApr 29, 2026
Performance & price
Output speed98 tok/s
Latency (TTFT)0.8s
First answer21s
Price in / out$1.5 / $7.5 per 1M
Blended price$3 / 1M
Coding index46.9

Median across API providers, via Artificial Analysis. Blended = 3:1 input:output. For reasoning models, latency includes thinking time. Methodology →

Capability over time

Append-only ledger — every observed change to this model's numbers.

API providers
Artificial Analysis$7.5/M