vs
5:8
overall capability wins (0 ties across 13 scored). Llama 4 Scout leads overall.
- Reasoning2:4
- Coding3:0
- Arena0:1
- Pricing0:3
- Speed0:2
- Specs0:3
Verdict
Llama 3.3 70B leads coding (widest gap: +4.9 on SWE-Rebench); Llama 4 Scout edges reasoning & knowledge; Llama 4 Scout ranks higher on LMArena (+4 Elo).
- Llama 3.3 70B leads coding (widest gap: +4.9 on SWE-Rebench)
- Llama 4 Scout edges reasoning & knowledge
- Llama 4 Scout ranks higher on LMArena (+4 Elo)
- Llama 4 Scout offers ~78.1x larger context (10M vs 128K)
- Llama 4 Scout allows ~4.0x more max output (33K vs 8K)
- Llama 4 Scout is ~3.1x cheaper on a blended token basis than Llama 3.3 70B
Cost for 1M in + 250K out
Illustrative chat workload at primary-provider list prices.
- Llama 3.3 70B
- $1.10
- Llama 4 Scout
- $0.33
Llama 4 Scout is about 3.4x cheaper on this workload.
Price per 1M tokens
Full comparison
Organization
Meta
Meta
Family
Llama / Muse
Llama / Muse
License
Llama 3.3 Community
Llama 4 Community
Open weights
Yes
Yes
Release
Dec 6, 2024
Apr 5, 2025
Knowledge cutoff
-
-
API / provider
Together / Fireworks (ref.)
Together / Fireworks (ref.)
Modalities
text → text
text, image → text
Specs
Context window
128K
10M
Max output
8K
33K
Parameters
70B
109B (17B act.)
Pricing
Input $/1M
$0.88
$0.18
Output $/1M
$0.88
$0.59
Blended $/1M (3∶1)
$0.88
$0.28
Speed
tok/s
80
110
TTFT (s)
0.35
0.3
Reasoning
MMLU-Pro
68.9
74.3
GPQA Diamond
49.8
58.7
Humanity's Last Exam
3.6
3.8
AIME 2025
—
14.0
MATH-500
—
84.4
Humanity's Last Exam (with tools)
—
—
AA-Omniscience Accuracy
19.0
15.2
AA-LCR v1.1
15.7
27.7
CritPt
—
—
MMMU-Pro
—
52.9
IFBench
47.1
39.5
Chartography
—
—
Chartography (With Tools)
—
—
Coding
SWE-bench Verified
—
9.1
SWE-bench Pro
—
—
SWE-bench Multilingual
—
—
LiveCodeBench
33.3
32.8
Terminal-Bench 2.1
4.9
3.7
Aider Polyglot
—
—
Terminal-Bench 3
—
—
BigCodeBench
46.9
—
SciCode
—
21.3
CursorBench
—
—
SWE-Rebench
11.5
6.6
NL2Repo-Bench
—
—
DeepSWE
—
—
WebDev Arena
—
—
Terminal-Bench 4.0
—
—
CursorBench 4.0
—
—
FrontierCode v1.1 (Main)
—
—
Arena
LMArena Elo
1,318
1,322
| Metric | Llama 3.3 70B | Llama 4 Scout | Delta |
|---|---|---|---|
| Identity | |||
| Organization | Meta | Meta | — |
| Family | Llama / Muse | Llama / Muse | — |
| License | Llama 3.3 Community | Llama 4 Community | — |
| Open weights | Yes | Yes | — |
| Release | Dec 6, 2024 | Apr 5, 2025 | — |
| Knowledge cutoff | - | - | — |
| API / provider | Together / Fireworks (ref.) | Together / Fireworks (ref.) | — |
| Modalities | text → text | text, image → text | — |
| Specs | |||
| Context window | 128K | 10M | B +9.9M |
| Max output | 8K | 33K | B +25K |
| Parameters | 70B | 109B (17B act.) | B +39 |
| Pricing | |||
| Input $/1M | $0.88 | $0.18 | B +$0.70 |
| Output $/1M | $0.88 | $0.59 | B +$0.29 |
| Blended $/1M (3∶1) | $0.88 | $0.28 | B +$0.60 |
| Speed | |||
| tok/s | 80 | 110 | B +30 |
| TTFT (s) | 0.35 | 0.3 | B +0.05 |
| Reasoning | |||
| MMLU-Pro | 68.9 | 74.3 | B +5.4 pts |
| GPQA Diamond | 49.8 | 58.7 | B +8.9 pts |
| Humanity's Last Exam | 3.6 | 3.8 | B +0.2 pts |
| AIME 2025 | — | 14.0 | — |
| MATH-500 | — | 84.4 | — |
| Humanity's Last Exam (with tools) | — | — | — |
| AA-Omniscience Accuracy | 19.0 | 15.2 | A +3.8 pts |
| AA-LCR v1.1 | 15.7 | 27.7 | B +12.0 pts |
| CritPt | — | — | — |
| MMMU-Pro | — | 52.9 | — |
| IFBench | 47.1 | 39.5 | A +7.6 pts |
| Chartography | — | — | — |
| Chartography (With Tools) | — | — | — |
| Coding | |||
| SWE-bench Verified | — | 9.1 | — |
| SWE-bench Pro | — | — | — |
| SWE-bench Multilingual | — | — | — |
| LiveCodeBench | 33.3 | 32.8 | A +0.5 pts |
| Terminal-Bench 2.1 | 4.9 | 3.7 | A +1.2 pts |
| Aider Polyglot | — | — | — |
| Terminal-Bench 3 | — | — | — |
| BigCodeBench | 46.9 | — | — |
| SciCode | — | 21.3 | — |
| CursorBench | — | — | — |
| SWE-Rebench | 11.5 | 6.6 | A +4.9 pts |
| NL2Repo-Bench | — | — | — |
| DeepSWE | — | — | — |
| WebDev Arena | — | — | — |
| Terminal-Bench 4.0 | — | — | — |
| CursorBench 4.0 | — | — | — |
| FrontierCode v1.1 (Main) | — | — | — |
| Arena | |||
| LMArena Elo | 1,318 | 1,322 | B +4 Elo |
Llama 3.3 70B
- Context
- 128K / 8K out
- Parameters
- 70B
- Price
- $0.88 / $0.88
- Speed
- 80 tok/s · 0.35s TTFT
- Modalities
- text → text
- License
- Llama 3.3 Community
Widely deployed open 70B instruct model; still a solid self-host baseline.
Llama 4 Scout
- Context
- 10M / 33K out
- Parameters
- 109B (17B act.)
- Price
- $0.18 / $0.59
- Speed
- 110 tok/s · 0.3s TTFT
- Modalities
- text, image → text
- License
- Llama 4 Community
Llama 4 variant with an extreme 10M context window for long-document retrieval workloads.
Benchmark charts
Winner bars are emphasized. Per-benchmark deltas sit above each chart.