vs
4:4
overall capability wins (0 ties across 8 scored). GPT-4 Turbo leads overall.
- Reasoning2:1
- Coding0:2
- Arena0:1
- Pricing3:0
- Speed2:0
- Specs2:0
Verdict
GPT-4 Turbo leads coding (widest gap: +6.0 on SciCode); GPT-4.1 Nano edges reasoning & knowledge; GPT-4.1 Nano offers ~7.8x larger context (1M vs 128K).
- GPT-4 Turbo leads coding (widest gap: +6.0 on SciCode)
- GPT-4.1 Nano edges reasoning & knowledge
- GPT-4.1 Nano offers ~7.8x larger context (1M vs 128K)
- GPT-4.1 Nano allows ~8.0x more max output (33K vs 4K)
- GPT-4.1 Nano is ~85.7x cheaper on a blended token basis than GPT-4 Turbo
Cost for 1M in + 250K out
Illustrative chat workload at primary-provider list prices.
- GPT-4.1 Nano
- $0.20
- GPT-4 Turbo
- $17.50
GPT-4.1 Nano is about 87.5x cheaper on this workload.
Price per 1M tokens
Full comparison
Organization
OpenAI
OpenAI
Family
GPT-4 / o-series
GPT-4 / o-series
License
Proprietary
Proprietary
Open weights
No
No
Release
Apr 14, 2025
Nov 6, 2023
Knowledge cutoff
-
-
API / provider
OpenAI
OpenAI
Modalities
text, image → text
text, image → text
Specs
Context window
1M
128K
Max output
33K
4K
Parameters
—
—
Pricing
Input $/1M
$0.10
$10
Output $/1M
$0.40
$30
Blended $/1M (3∶1)
$0.18
$15
Speed
tok/s
160
55
TTFT (s)
0.18
0.6
Reasoning
MMLU-Pro
65.7
69.4
GPQA Diamond
51.2
—
Humanity's Last Exam
3.8
3.1
AIME 2025
24.0
—
MATH-500
84.8
73.7
Humanity's Last Exam (with tools)
—
—
AA-Omniscience Accuracy
13.7
—
AA-LCR v1.1
20.3
—
CritPt
—
—
MMMU-Pro
40.1
—
IFBench
32.0
—
Chartography
—
—
Chartography (With Tools)
—
—
Coding
SWE-bench Verified
—
—
SWE-bench Pro
—
—
SWE-bench Multilingual
—
—
LiveCodeBench
32.6
37.3
Terminal-Bench 2.1
3.7
—
Aider Polyglot
8.9
—
Terminal-Bench 3
—
—
BigCodeBench
—
48.2
SciCode
25.9
31.9
CursorBench
—
—
SWE-Rebench
0.3
—
NL2Repo-Bench
—
—
DeepSWE
—
—
WebDev Arena
—
—
Terminal-Bench 4.0
—
—
CursorBench 4.0
—
—
FrontierCode v1.1 (Main)
—
—
Arena
LMArena Elo
1,322
1,324
| Metric | GPT-4.1 Nano | GPT-4 Turbo | Delta |
|---|---|---|---|
| Identity | |||
| Organization | OpenAI | OpenAI | — |
| Family | GPT-4 / o-series | GPT-4 / o-series | — |
| License | Proprietary | Proprietary | — |
| Open weights | No | No | — |
| Release | Apr 14, 2025 | Nov 6, 2023 | — |
| Knowledge cutoff | - | - | — |
| API / provider | OpenAI | OpenAI | — |
| Modalities | text, image → text | text, image → text | — |
| Specs | |||
| Context window | 1M | 128K | A +872K |
| Max output | 33K | 4K | A +29K |
| Parameters | — | — | — |
| Pricing | |||
| Input $/1M | $0.10 | $10 | A +$9.90 |
| Output $/1M | $0.40 | $30 | A +$29.60 |
| Blended $/1M (3∶1) | $0.18 | $15 | A +$14.82 |
| Speed | |||
| tok/s | 160 | 55 | A +105 |
| TTFT (s) | 0.18 | 0.6 | A +0.42 |
| Reasoning | |||
| MMLU-Pro | 65.7 | 69.4 | B +3.7 pts |
| GPQA Diamond | 51.2 | — | — |
| Humanity's Last Exam | 3.8 | 3.1 | A +0.7 pts |
| AIME 2025 | 24.0 | — | — |
| MATH-500 | 84.8 | 73.7 | A +11.1 pts |
| Humanity's Last Exam (with tools) | — | — | — |
| AA-Omniscience Accuracy | 13.7 | — | — |
| AA-LCR v1.1 | 20.3 | — | — |
| CritPt | — | — | — |
| MMMU-Pro | 40.1 | — | — |
| IFBench | 32.0 | — | — |
| Chartography | — | — | — |
| Chartography (With Tools) | — | — | — |
| Coding | |||
| SWE-bench Verified | — | — | — |
| SWE-bench Pro | — | — | — |
| SWE-bench Multilingual | — | — | — |
| LiveCodeBench | 32.6 | 37.3 | B +4.7 pts |
| Terminal-Bench 2.1 | 3.7 | — | — |
| Aider Polyglot | 8.9 | — | — |
| Terminal-Bench 3 | — | — | — |
| BigCodeBench | — | 48.2 | — |
| SciCode | 25.9 | 31.9 | B +6.0 pts |
| CursorBench | — | — | — |
| SWE-Rebench | 0.3 | — | — |
| NL2Repo-Bench | — | — | — |
| DeepSWE | — | — | — |
| WebDev Arena | — | — | — |
| Terminal-Bench 4.0 | — | — | — |
| CursorBench 4.0 | — | — | — |
| FrontierCode v1.1 (Main) | — | — | — |
| Arena | |||
| LMArena Elo | 1,322 | 1,324 | B +2 Elo |
GPT-4.1 Nano
- Context
- 1M / 33K out
- Parameters
- —
- Price
- $0.10 / $0.40
- Speed
- 160 tok/s · 0.18s TTFT
- Modalities
- text, image → text
- License
- Proprietary
Cheapest GPT-4.1 tier for classification, routing, and high-volume extraction with a full 1M context window.
GPT-4 Turbo
- Context
- 128K / 4K out
- Parameters
- —
- Price
- $10 / $30
- Speed
- 55 tok/s · 0.6s TTFT
- Modalities
- text, image → text
- License
- Proprietary
128K-context GPT-4 refresh with vision — long-running production workhorse before GPT-4o / 4.1.
Benchmark charts
Winner bars are emphasized. Per-benchmark deltas sit above each chart.