vs
6:1
overall capability wins (0 ties across 7 scored). Mistral Medium 3.5 leads overall.
- Reasoning2:0
- Coding1:0
- Arena1:0
- Pricing0:3
- Speed2:0
- Specs2:1
Verdict
Mistral Medium 3.5 leads coding (widest gap: +21.4 on SciCode); Mistral Medium 3.5 ranks higher on LMArena (+197 Elo); Mistral Medium 3.5 offers ~3.9x larger context (256K vs 66K).
- Mistral Medium 3.5 leads coding (widest gap: +21.4 on SciCode)
- Mistral Medium 3.5 ranks higher on LMArena (+197 Elo)
- Mistral Medium 3.5 offers ~3.9x larger context (256K vs 66K)
- Mistral Medium 3.5 allows ~7.8x more max output (64K vs 8K)
- Mixtral 8x22B is ~2.5x cheaper on a blended token basis than Mistral Medium 3.5
Cost for 1M in + 250K out
Illustrative chat workload at primary-provider list prices.
- Mistral Medium 3.5
- $3.38
- Mixtral 8x22B
- $1.50
Mixtral 8x22B is about 2.3x cheaper on this workload.
Price per 1M tokens
Full comparison
Organization
Mistral
Mistral
Family
Mistral
Mistral
License
Modified MIT
Apache 2.0
Open weights
Yes
Yes
Release
Apr 22, 2026
Apr 17, 2024
Knowledge cutoff
-
-
API / provider
Mistral
Together / Fireworks (ref.)
Modalities
text, image → text
text → text
Specs
Context window
256K
66K
Max output
64K
8K
Parameters
128B
141B (39B act.)
Pricing
Input $/1M
$1.50
$1.20
Output $/1M
$7.50
$1.20
Blended $/1M (3∶1)
$3
$1.20
Speed
tok/s
80
50
TTFT (s)
0.4
0.5
Reasoning
MMLU-Pro
—
53.7
GPQA Diamond
74.8
33.2
Humanity's Last Exam
13.8
4.0
AIME 2025
—
—
MATH-500
—
54.5
Humanity's Last Exam (with tools)
—
—
AA-Omniscience Accuracy
24.7
—
AA-LCR v1.1
69.3
—
CritPt
—
—
MMMU-Pro
64.9
—
IFBench
68.8
—
Chartography
—
—
Chartography (With Tools)
—
—
Coding
SWE-bench Verified
—
—
SWE-bench Pro
—
—
SWE-bench Multilingual
—
—
LiveCodeBench
—
14.8
Terminal-Bench 2.1
50.6
—
Aider Polyglot
—
—
Terminal-Bench 3
—
—
BigCodeBench
—
40.6
SciCode
40.2
18.8
CursorBench
—
—
SWE-Rebench
—
—
NL2Repo-Bench
—
—
DeepSWE
—
—
WebDev Arena
1,265
—
Terminal-Bench 4.0
—
—
CursorBench 4.0
—
—
FrontierCode v1.1 (Main)
—
—
Arena
LMArena Elo
1,426
1,229
| Metric | Mistral Medium 3.5 | Mixtral 8x22B | Delta |
|---|---|---|---|
| Identity | |||
| Organization | Mistral | Mistral | — |
| Family | Mistral | Mistral | — |
| License | Modified MIT | Apache 2.0 | — |
| Open weights | Yes | Yes | — |
| Release | Apr 22, 2026 | Apr 17, 2024 | — |
| Knowledge cutoff | - | - | — |
| API / provider | Mistral | Together / Fireworks (ref.) | — |
| Modalities | text, image → text | text → text | — |
| Specs | |||
| Context window | 256K | 66K | A +190K |
| Max output | 64K | 8K | A +56K |
| Parameters | 128B | 141B (39B act.) | B +13 |
| Pricing | |||
| Input $/1M | $1.50 | $1.20 | B +$0.30 |
| Output $/1M | $7.50 | $1.20 | B +$6.30 |
| Blended $/1M (3∶1) | $3 | $1.20 | B +$1.80 |
| Speed | |||
| tok/s | 80 | 50 | A +30 |
| TTFT (s) | 0.4 | 0.5 | A +0.10 |
| Reasoning | |||
| MMLU-Pro | — | 53.7 | — |
| GPQA Diamond | 74.8 | 33.2 | A +41.6 pts |
| Humanity's Last Exam | 13.8 | 4.0 | A +9.8 pts |
| AIME 2025 | — | — | — |
| MATH-500 | — | 54.5 | — |
| Humanity's Last Exam (with tools) | — | — | — |
| AA-Omniscience Accuracy | 24.7 | — | — |
| AA-LCR v1.1 | 69.3 | — | — |
| CritPt | — | — | — |
| MMMU-Pro | 64.9 | — | — |
| IFBench | 68.8 | — | — |
| Chartography | — | — | — |
| Chartography (With Tools) | — | — | — |
| Coding | |||
| SWE-bench Verified | — | — | — |
| SWE-bench Pro | — | — | — |
| SWE-bench Multilingual | — | — | — |
| LiveCodeBench | — | 14.8 | — |
| Terminal-Bench 2.1 | 50.6 | — | — |
| Aider Polyglot | — | — | — |
| Terminal-Bench 3 | — | — | — |
| BigCodeBench | — | 40.6 | — |
| SciCode | 40.2 | 18.8 | A +21.4 pts |
| CursorBench | — | — | — |
| SWE-Rebench | — | — | — |
| NL2Repo-Bench | — | — | — |
| DeepSWE | — | — | — |
| WebDev Arena | 1,265 | — | — |
| Terminal-Bench 4.0 | — | — | — |
| CursorBench 4.0 | — | — | — |
| FrontierCode v1.1 (Main) | — | — | — |
| Arena | |||
| LMArena Elo | 1,426 | 1,229 | A +197 Elo |
Mistral Medium 3.5
- Context
- 256K / 64K out
- Parameters
- 128B
- Price
- $1.50 / $7.50
- Speed
- 80 tok/s · 0.4s TTFT
- Modalities
- text, image → text
- License
- Modified MIT
Mistral's April 2026 frontier multimodal model (open weights, Modified MIT) powering Vibe remote agents at $1.50/$7.50.
Mixtral 8x22B
- Context
- 66K / 8K out
- Parameters
- 141B (39B act.)
- Price
- $1.20 / $1.20
- Speed
- 50 tok/s · 0.5s TTFT
- Modalities
- text → text
- License
- Apache 2.0
Larger Mixtral MoE (141B / 39B active) — strong open chat and coding before Mistral Large 2.
Benchmark charts
Winner bars are emphasized. Per-benchmark deltas sit above each chart.