Data points: 95
Model compare
The readout for Claude 3 Haiku and GPT-5.2, before the detailed comparison sheet.
Weighted outcome: GPT-5.2. Benchmark capability categories carry 80%, while price, API performance, and availability carry 20%.
Decision read
GPT-5.2
GPT-5.2 has the higher weighted result; Model A / B score 0 to 100.
Evidence depth
95 data points
Includes 16 benchmark rows, 0 audit samples, and 6 provider examples.
Selection signal
Start with GPT-5.2
The charts below split 22 high-signal samples across speed, scores, and audit health.
Switch either side of this report to compare another model with the same LMSpeed data pipeline.
Select a different model to open a new comparison URL.
This report only uses LMSpeed data for Claude 3 Haiku and GPT-5.2: pricing, speed aggregates, third-party benchmark scores, and shared provider samples.
| Model compare | Claude 3 Haiku | GPT-5.2 |
|---|---|---|
| Overall leader | Contender | Leading |
| Weighted overall score | 0.0 pts | 100.0 pts |
| Benchmark category leads | 0 categories | 6 categories |
| Operational advantages | No data | Cheapest input price, Free providers, Provider coverage |
| Context window | 200K tokens | 400K tokens |
| Max output | 4.1K tokens | 128K tokens |
| Modalities | Input TextImage Output Text |
The overall result weights benchmark capability categories at 80% and price, API speed/latency, and availability at 20%. Recent test volume does not affect the winner, and missing benchmark categories are excluded.
| Model compare | Claude 3 Haiku | GPT-5.2 |
|---|---|---|
| Developer | Anthropic | OpenAI |
| Released | Mar 2024 | Dec 2025 |
| Parameters | No data | No data |
| Tokenizer | Claude | GPT |
| Knowledge cutoff | 2023-08-31 | No data |
| OpenRouter ID | anthropic/claude-3-haiku | openai/gpt-5.2 |
| References | No data | No data |
This report only uses LMSpeed data for Claude 3 Haiku and GPT-5.2: pricing, speed aggregates, third-party benchmark scores, and shared provider samples.
Claude 3 Haiku
Claude 3 Haiku does not clearly lead in the benchmark or operational dimensions shared by both models.
GPT-5.2
GPT-5.2 is stronger in benchmark categories (Agents, Coding, Reasoning, Knowledge, Math) and operational dimensions (Cheapest input price, Free providers, Provider coverage).
Third-party benchmark profile synced into LMSpeed; only metrics available for both models are shown.
Compare benchmark category scores on a 0-100 scale. Select a category to inspect the gap.
Avg. score
Claude 3 Haiku
37.5
Avg. score
GPT-5.2
52.8
GPT-5.2 leads by 10.2
GPT-5.2 leads by 24.0
GPT-5.2 leads by 19.1
GPT-5.2 leads by 16.4
GPT-5.2 leads by 19.7
No data
GPT-5.2
GPT-5.2 leads by 12.8
Metric-level scores with benchmark source, rank depth, confidence, error, and evaluation date where available.
Claude 3 Haiku
winner$0.250/M
Rank #36/162 · confidence 4
GPT-5.2
$1.75/M
Rank #120/162 · confidence 4
Claude 3 Haiku
winner$0.500/M
Rank #40/162 · confidence 4
GPT-5.2
$4.81/M
Rank #132/162 · confidence 4
Claude 3 Haiku
winner$1.25/M
Rank #50/162 · confidence 4
GPT-5.2
$14.00/M
Rank #133/162 · confidence 4
Claude 3 Haiku
21.0
Rank #87/89 · confidence 1 · eval date 2024-03-01
GPT-5.2
winner72.7
Rank #13/89 · confidence 3 · eval date 2025-12-11
Claude 3 Haiku
3.9%
Rank #165/187 · confidence 4
GPT-5.2
winner7.3%
Rank #109/187 · confidence 4
Claude 3 Haiku
37.4%
Rank #180/188 · confidence 4
GPT-5.2
winner71.2%
Rank #110/188 · confidence 4
Claude 3 Haiku
18.6
Rank #95/96 · confidence 1 · eval date 2024-03-01
GPT-5.2
winner52.1
Rank #18/96 · confidence 3 · eval date 2025-12-11
Claude 3 Haiku
15.4%
Rank #111/115 · confidence 4
GPT-5.2
winner66.9%
Rank #36/115 · confidence 4
Claude 3 Haiku
18.6%
Rank #177/185 · confidence 4
GPT-5.2
winner40.4%
Rank #64/185 · confidence 4
Claude 3 Haiku
17.2
Rank #81/90 · confidence 1 · eval date 2024-03-01
GPT-5.2
winner43.8
Rank #16/90 · confidence 3 · eval date 2025-12-11
Claude 3 Haiku
37.4
Rank #96/96 · confidence 1 · eval date 2024-03-01
GPT-5.2
winner90.3
Rank #19/96 · confidence 3 · eval date 2025-12-11
Claude 3 Haiku
3.9
Rank #91/96 · confidence 1 · eval date 2024-03-01
GPT-5.2
winner35.4
Rank #21/96 · confidence 3 · eval date 2025-12-11
Claude 3 Haiku
3.9
Rank #99/99 · confidence 1 · eval date 2024-03-01
GPT-5.2
winner42.2
Rank #22/99 · confidence 3 · eval date 2025-12-11
Claude 3 Haiku
78.2
Rank #43/90 · confidence 1 · eval date 2024-03-01
GPT-5.2
winner79.7
Rank #39/90 · confidence 3 · eval date 2025-12-11
Claude 3 Haiku
36.1
Rank #80/86 · confidence 1 · eval date 2024-03-01
GPT-5.2
winner75.4
Rank #21/86 · confidence 3 · eval date 2025-12-11
Claude 3 Haiku
21.1
Rank #79/84 · confidence 1 · eval date 2024-03-01
GPT-5.2
winner84.8
Rank #44/84 · confidence 3 · eval date 2025-12-11
Latest completed audits from shared providers, with four safety and integrity score groups plus report links.
| Provider | Claude 3 Haiku | GPT-5.2 |
|---|---|---|
| No completed audits are available from shared providers yet. | ||
Speed aggregates and input/output pricing share each provider row for real API selection and migration cost checks.
| Provider | Claude 3 Haiku | GPT-5.2 |
|---|---|---|
5 tests | Claude 3 Haiku claude-3-haiku-20240307 speed / latency N/A / N/A input / output $0.030/M/$0.152/M | GPT-5.2 gpt-5.2-chat speed / latency 64 tok/s / 5256ms input / output $0.212/M/$1.70/M |
0 tests | Claude 3 Haiku speed / latency N/A / N/A input / output No data | GPT-5.2 speed / latency N/A / N/A input / output No data |
0 tests | Claude 3 Haiku speed / latency N/A / N/A input / output No data | GPT-5.2 speed / latency N/A / N/A input / output No data |
0 tests | Claude 3 Haiku speed / latency N/A / N/A input / output No data | GPT-5.2 speed / latency N/A / N/A input / output No data |
0 tests | Claude 3 Haiku speed / latency N/A / N/A input / output No data | GPT-5.2 speed / latency N/A / N/A input / output No data |
Claude 3 Haiku claude-3-haiku-20240307 speed / latency No data input / output $0.154/M/$0.771/M | GPT-5.2 gpt-5.2-chat-2026-02-10 speed / latency No data input / output $56.51/M |
Weighted outcome: GPT-5.2. Benchmark capability categories carry 80%, while price, API performance, and availability carry 20%.
Input
Output
| Features | Text inputImage inputText outputTool calling | Text inputImage inputFile inputText outputTool callingStructured outputsJSON modeReasoning |
|---|
