OpenAI, proprietary
GPT-4o mini
GPT-4o mini by OpenAI ranks 343rd of 354 ranked models on the Noometry Index as of October 2026, with a score of 25.5. Its strongest category is agentic & tool use, where it ranks 101st. API pricing starts at $0.15 per million input tokens and $0.60 per million output tokens, with a 128K-token context window.
Last verified
Specifications
- Noometry rank
- #343 of 354
- Index score
- 25.5
- Evidence
- Confirmed 60 results
- Provider
- OpenAI
- Released
- July 18, 2024
- Weights
- Proprietary
- Reasoning
- No
- Context window
- 128K
- Max output
- 16K
- Input price
- $0.15 / M
- Output price
- $0.60 / M
- Blended price
- $0.26 / M
- Output speed
- 120 tokens/s Kagi
- Value
- #63 of 219
- Knowledge cutoff
- September 2023
- Input
- text, image, pdf
Category scores
Each category score combines every public result we have in that category.
- Coding 22.0
- Agentic & Tool Use 27.5
- Reasoning 8.7
- Math 10.4
- Knowledge 17.7
- Multimodal 25.9
- Multilingual 42.0
- Instruction Following 61.9
- Long Context 39.1
- Writing & Preference 39.5
| Category | Score | Rank | Results |
|---|---|---|---|
| Coding | 22.0 | #335 | 6 |
| Agentic & Tool Use | 27.5 | #101 | 1 |
| Reasoning | 8.7 | #347 | 10 |
| Math | 10.4 | #314 | 6 |
| Knowledge | 17.7 | #284 | 6 |
| Multimodal | 25.9 | #122 | 4 |
| Multilingual | 42.0 | #199 | 1 |
| Instruction Following | 61.9 | #239 | 3 |
| Long Context | 39.1 | #186 | 1 |
| Writing & Preference | 39.5 | #248 | 7 |
Strengths and weaknesses
Categories where GPT-4o mini places highest and lowest among the models ranked in each, with its score against that category's median.
Strongest categories
| Category | Score | vs median | Rank |
|---|---|---|---|
| Long Context | 39.1 | −1.8 | #186 of 296, top 63% |
| Agentic & Tool Use | 27.5 | −2.9 | #101 of 154, top 66% |
| Multilingual | 42.0 | −5.4 | #199 of 297, top 68% |
Closest competitors
The models ranked just above and below GPT-4o mini. When scores are this close, price and speed are often the better way to choose.
| Model | Rank | Score | Blended $/M | Speed | |
|---|---|---|---|---|---|
| DeepSeek-R1-Distill-Qwen-1.5B | #339 | 26.1 | — | — | Compare |
| Claude 3 Haiku | #340 | 25.9 | — | 41 | Compare |
| Gemma 2 9B | #341 | 25.9 | — | — | Compare |
| Dolly 2.0-12b | #342 | 25.5 | — | — | Compare |
| Llama 3-8B | #344 | 25.5 | — | — | Compare |
| Claude 2.1 | #345 | 25.2 | — | — | Compare |
| Claude 2 | #346 | 25.0 | — | — | Compare |
| DeepSeek LLM 67B | #347 | 24.9 | — | — | Compare |
Sponsored placements are available on pages like this one. Advertise on Noometry
Benchmark results
Every published result we track, with its source. Bold rows are the ones used for ranking; where several exist we prefer independent runs over self-reported numbers.
Coding
| Benchmark | Score | Position | Setting | Source | Date |
|---|---|---|---|---|---|
| Aider Polyglot | 3.6% | #44 of 44, top 100% | Epoch AI | ||
| WeirdML | 11.8% | #111 of 119, top 94% | Epoch AI | ||
| BigCodeBench Instruct | 46.1% | #13 of 64, top 21% | BigCodeBench | 2024-07-18 | |
| LiveBench Coding | 43.1% | #22 of 39, top 57% | Epoch AI | ||
| LMArena Coding | 1290 | #205 of 294, top 70% | LMArena | 2026-10-08 | |
| BigCodeBench Complete | 57.4% | #14 of 66, top 22% | BigCodeBench | 2024-07-18 | |
| HumanEval+ | 83.5% | #8 of 45, top 18% | july 2024 | EvalPlus | |
| MBPP+ | 72.2% | #12 of 38, top 32% | july 2024 | EvalPlus |
Agentic & Tool Use
| Benchmark | Score | Position | Setting | Source | Date |
|---|---|---|---|---|---|
| BALROG | 17.4% | #27 of 35, top 78% | Epoch AI |
Reasoning
| Benchmark | Score | Position | Setting | Source | Date |
|---|---|---|---|---|---|
| ARC-AGI-2 | 0% | #78 of 83, top 94% | Epoch AI | ||
| SimpleBench | 10.7% | #77 of 77, top 100% | Epoch AI | ||
| Kagi LLM Benchmark | 28.8% | #94 of 99, top 95% | Kagi LLM Benchmark | ||
| Chess Puzzles | 0% | #116 of 129, top 90% | Epoch AI | 2026-07-15 | |
| LiveBench Reasoning | 32.8% | #29 of 39, top 75% | Epoch AI | ||
| LMArena Hard Prompts | 1267 | #208 of 297, top 71% | LMArena | 2026-10-08 | |
| Mystery Game Puzzles | 12% | #57 of 74, top 78% | Epoch AI | 2026-08-27 | |
| DTBench | 54.4% | #124 of 151, top 83% | Epoch AI | ||
| LiveBench Data Analysis | 50% | #23 of 39, top 59% | Epoch AI | ||
| LMCA | 10.4% | #110 of 125, top 88% | Epoch AI | ||
| Epoch Capabilities Index | 126.56 | #153 of 213, top 72% | Epoch AI | 2024-07-18 | |
| LiveBench | 41.3% | #30 of 39, top 77% | Epoch AI | ||
| PIQA | 88.7% | Best of 27 | Epoch AI |
Math
| Benchmark | Score | Position | Setting | Source | Date |
|---|---|---|---|---|---|
| FrontierMath (Tiers 1-3) | 0.7% | #79 of 81, top 98% | Epoch AI | 2026-08-27 | |
| OTIS Mock AIME 2024-2025 | 6.9% | #141 of 173, top 82% | Epoch AI | 2025-07-30 | |
| Omni-MATH | 28% | #42 of 57, top 74% | HELM Capabilities | ||
| LiveBench Math | 36.3% | #30 of 39, top 77% | Epoch AI | ||
| LMArena Math | 1267 | #205 of 285, top 72% | LMArena | 2026-10-08 | |
| MATH Level 5 | 52.6% | #44 of 79, top 56% | Epoch AI | 2025-01-27 | |
| GSM8K | 91.3% | #4 of 38, top 11% | Epoch AI |
Knowledge
| Benchmark | Score | Position | Setting | Source | Date |
|---|---|---|---|---|---|
| GPQA Diamond | 37.7% | #154 of 186, top 83% | Epoch AI | 2025-01-27 | |
| SimpleQA Verified | 8.3% | #76 of 77, top 99% | Epoch AI | 2026-08-31 | |
| MMLU-Pro | 60.3% | #43 of 58, top 75% | HELM Capabilities | ||
| Confabulations (lower is better) | 37.2% | #50 of 51, top 99% | Lech Mazur benchmarks | ||
| GPQA (HELM) | 36.8% | #47 of 57, top 83% | HELM Capabilities | ||
| LMArena Expert | 1235 | #206 of 273, top 76% | LMArena | 2026-10-08 | |
| BoolQ | 88.7% | #2 of 23, top 9% | Epoch AI | ||
| MMLU | 81.8% | #13 of 81, top 17% | Epoch AI |
Multimodal
| Benchmark | Score | Position | Setting | Source | Date |
|---|---|---|---|---|---|
| LMArena Vision | 1066 | #112 of 122, top 92% | LMArena | 2026-10-09 | |
| Video-MME | 64.8% | #9 of 15, top 60% | Epoch AI | ||
| GeoBench | 64% | #14 of 25, top 57% | Epoch AI | ||
| VPCT | 34% | #21 of 24, top 88% | Epoch AI |
Multilingual
| Benchmark | Score | Position | Setting | Source | Date |
|---|---|---|---|---|---|
| LMArena Non-English | 1266 | #199 of 297, top 68% | LMArena | 2026-10-08 | |
| LMArena Chinese | 1265 | #201 of 285, top 71% | LMArena | 2026-10-08 | |
| LMArena French | 1297 | #163 of 223, top 74% | LMArena | 2026-10-08 | |
| LMArena German | 1272 | #164 of 231, top 71% | LMArena | 2026-10-08 | |
| LMArena Japanese | 1216 | #150 of 211, top 72% | LMArena | 2026-10-08 | |
| LMArena Korean | 1195 | #163 of 213, top 77% | LMArena | 2026-10-08 | |
| LMArena Russian | 1275 | #193 of 283, top 69% | LMArena | 2026-10-08 | |
| LMArena Spanish | 1276 | #171 of 226, top 76% | LMArena | 2026-10-08 |
Instruction Following
| Benchmark | Score | Position | Setting | Source | Date |
|---|---|---|---|---|---|
| LiveBench Instruction Following | 56.8% | #32 of 39, top 83% | Epoch AI | ||
| IFEval | 78.2% | #46 of 57, top 81% | HELM Capabilities | ||
| LMArena Instruction Following | 1258 | #204 of 298, top 69% | LMArena | 2026-10-08 |
Long Context
| Benchmark | Score | Position | Setting | Source | Date |
|---|---|---|---|---|---|
| LMArena Longer Query | 1289 | #197 of 291, top 68% | LMArena | 2026-10-08 |
Writing & Preference
| Benchmark | Score | Position | Setting | Source | Date |
|---|---|---|---|---|---|
| LMArena Text | 1286 | #203 of 297, top 69% | LMArena | 2026-10-08 | |
| LMArena Creative Writing | 1268 | #193 of 295, top 66% | LMArena | 2026-10-08 | |
| Short-Story Creative Writing | 67.2% | #33 of 39, top 85% | Epoch AI | ||
| EQ-Bench Creative Writing | 873 | #100 of 115, top 87% | EQ-Bench | ||
| WildBench | 79.1% | #35 of 57, top 62% | HELM Capabilities | ||
| LMArena Multi-Turn | 1285 | #197 of 295, top 67% | LMArena | 2026-10-08 | |
| LiveBench Language | 28.6% | #28 of 39, top 72% | Epoch AI |
API pricing by provider
| Route | Input $/M | Output $/M | Cached input $/M | Checked |
|---|---|---|---|---|
| azure | $0.15 | $0.60 | $0.075 | 2026-10-10 |
| openai | $0.15 | $0.60 | $0.075 | 2026-10-10 |
| openrouter | $0.15 | $0.60 | $0.075 | 2026-10-10 |
Compare GPT-4o mini
- GPT-4o mini vs Dolly 2.0-12b
- GPT-4o mini vs Llama 3-8B
- GPT-4o mini vs Gemma 2 9B
- GPT-4o mini vs Claude 2.1
- GPT-4o mini vs Claude 3 Haiku
- GPT-4o mini vs Claude 2
- GPT-4o mini vs Claude Fable 5.1
- GPT-4o mini vs Gemini 3.8 Flash
- GPT-4o mini vs Kimi K3
- GPT-4o mini vs Grok 4.6
- GPT-4o mini vs Qwen3.8 Max
- GPT-4o mini vs GLM-5.3
- GPT-4o mini vs Muse Spark 1.3
- GPT-4o mini vs DeepSeek V4 Pro
Other OpenAI models
- GPT-6 Astra70.8
- GPT-6.1 Sol65.6
- GPT-5.6 Sol65.0
- GPT-5.5 Pro64.3
- GPT-5.563.4
- GPT-6 Sol61.8
- GPT-5.459.4
- GPT-5.6 Terra59.2
Frequently asked questions
How good is GPT-4o mini?
GPT-4o mini by OpenAI ranks 343rd of 354 ranked models on the Noometry Index as of October 2026, with a score of 25.5. Its strongest category is agentic & tool use, where it ranks 101st. API pricing starts at $0.15 per million input tokens and $0.60 per million output tokens, with a 128K-token context window.
How much does GPT-4o mini cost?
GPT-4o mini costs $0.15 per million input tokens and $0.60 per million output tokens on OpenAI's own API, with cached input at $0.075.
What is GPT-4o mini's context window?
GPT-4o mini accepts up to 128K tokens of input and can write up to 16K tokens in one response.
Is GPT-4o mini open source?
No. GPT-4o mini is proprietary and available only through OpenAI's API and partner platforms.
How fast is GPT-4o mini?
GPT-4o mini generated about 120 output tokens per second in the Kagi LLM Benchmark's timed runs. Speed varies by provider, load and reasoning effort.
What are GPT-4o mini's strengths and weaknesses?
Relative to other ranked models, GPT-4o mini places best in long context, agentic & tool use, multilingual and lowest in reasoning, coding, math.
What is GPT-4o mini best at?
Its best category is agentic & tool use, where it ranks 101st on Noometry.